Namihei Odaira was born on the 15th of January in 1874 in Ienaka, Shimotsuga, Tochigi Prefecture, Empire of Japan to Sōhachi and Chiyo Odaira. This was the Meiji era, and at the time of Odaira’s birth, the Japanese Industrial Revolution was well underway and Japan’s first hydroelectric generator was completed in Kyoto. In keeping with the times, Odaira’s parents insisted that their children be well educated, and that, as a matter of business practicality, their children learn English. Around the age of sixteen, Odaira’s father fell ill and passed away. Odaira’s older brother then left school to support the family and to ensure that Odaira could continue his studies.

As he was completing his secondary education, Odaira felt anxious about his future. What should he study? At the age of twenty two, Odaira enrolled in Tokyo Imperial University’s electrical engineering department. He would have graduated after concluding his third year, but being young, he became absorbed by hobbies (photography in particular), and he had to repeat a year. He graduated at the age of twenty six in 1900.

The rise in electrification around the world meant a rise in copper mining, and those copper mines needed electricity themselves. Odaira was hired by Fujita & Co and assigned to the Kosaka Mine in Akita under the direction of mine chief Fusanosuke Kuhara. His first job was to serve as the engineering section chief, but shortly after he started in that position, Kuhara set him to work on a power station for the mine. He completed this task in just two years, and he was now convinced that electricity was the future of Japan. He left the Fujita zaibatsu and went to work for Tokyo Dento. This job had him working on power transmission equipment.

Kuhara contacted Odaira sometime in late 1903 or early 1904, and informed him Odaira of his intention to acquire the Akazawa copper mine in Hitachi City in Ibaraki. Kuhara wanted to avail himself of Odaira’s expertise and capabilities. Odaira accepted the job as chief of engineering, completed the Nakazato power station shortly after joining, and then went to work overhauling of the Akazawa mine’s machinery and facilities. In 1905, the Akazawa mine was renamed the Hitachi mine, and sometime shortly after Kuhara started the Kuhara zaibatsu and moved his various enterprises into it.

Largely due to the work of Odaira and his engineering team, production at the mine increased, and this brought about the need for even more power. The result was the design and construction of the 4000 kW Ishioka power station.

Odaira was likely a little obsessive. While working incredible hours for his employer, he and his team were also working on designs for electrical equipment. The first product of their research and design efforts was a 5hp/3.7kW electric motor completed in 1910. He took this to Kuhara, there was a brief debate, and Odaira was permitted to pursue the idea. Offices were finished that November, and a team of about four hundred was assembled. He chose to use the name Hitachi 日立 for this venture, combining the kanji for sun and rise (fitting for a company in the land of the rising sun), and it was a natural fit given the name of the city and mine.

Initial production was difficult, defect rates were high, and this spurred Hitachi to start an in-house training school. This helped, but after setting up a large generator for a customer resulted in another failure, Odaira established a formal equipment testing department, and a formal research and development department.

The company moved to Tokyo in 1917 only to burn down two years later. Despite the loss, this allowed Hitachi to become fully independent of Kuhara zaibatsu (which would later become Nissan) in 1920 as Hitachi, Ltd with Odaira as the managing director. On the 1st of September in 1923, an earthquake damaged Tokyo’s power grid and generation systems. Hitachi’s work in turning this around gained them social credibility and name recognition. The following year, Hitachi completed the ED15 electric locomotive, which combined with their recently received reputation, spurred orders from the Japanese government. Electric fans followed in 1926 and became the company’s first export. Those exports, however, were short lived with the arrival of the Great Depression in 1929. Odaira became the president of the company that year. In 1932, the company began producing elevators, and refrigerators. In 1939, the company built water turbines. As with most companies in countries involved in WWII, Hitachi was required to build for the war effort. This made their factories prime targets during the war. Odaira left the company in 1947 by direction of the US occupation. After the war, the company rebuilt yet again. In 1949, the company began producing construction machinery greatly adding to the country’s reconstruction efforts.

While Hitachi operated in many fields (electrical generation, consumer electronics, locomotives, nuclear reactors, auto parts), the rest of this article will focus on (mostly) on the company’s computing efforts.

Hitachi began research into computing in 1951. The earliest work was in analog computers, and digital computing research began in 1956 with parametron logic. Why would a company that dealt in electrical power generation, construction, appliances, and mining be researching computers in the 1950s? Well, for Hitachi, knowing the exact sag and tension of power lines was rather important. For these calculations, the company completed the HIPAC MK-1 in 1957. The HIPAC MK-1 (HItachi Parametron Automatic Computer) was loosely based upon the EDSAC (and the ILLIAC to a lesser extent), but it was quite obviously different and original as it made use of parametrons. The MK-1 used magnetic drum memory, paper tape input (also used for system bootstrap), and could output to paper tape or a printer. The MK-1 performed its duties well and saved Hitachi quite a bit of time.

The MK-1 was good. Hitachi then enhanced it, provided standardized input, output, and housing for it, and displayed it at a computer exhibition in Paris in June of 1959. This was the HIPAC 101. Shipments of the machine began in July of 1960. An improved version, the 103, shipped in December of 1961. The updated HIPAC utilized core memory, 48bit words, with the capability to address 8192 words. It was exceedingly CISCy with 110 different types of instructions. It was capable of hardware floating point, and it shipped with FORTRAN. This was the final parametron computer from Hitachi. The most impressive bit, to me, reading about the HIPAC is that the mean time to failure was high. Even with the MK-1, it would seem that it ran for a solid seven hours for a single calculation. This is the kind of figure expected of fully transistorized machines, and even more particularly of machines that employed integrated components.

In the mid-1950s, a team at the Japanese National Railways’ Railway Technical Research Institute led by Mamoru Hosaka and Yutaka Ohno began designing logic for a computer that could handle seat reservations on a locomotive. This design was completed in 1958 and then given to Hitachi to implement. The machine, MARS-1 (magnetic-electronic automatic reservation system), was completed in July of 1959 at Hitachi’s Totsuka Plant by a team led by Yasuhiko Tani.

The MARS-1 was unique. The logic was implemented with 4500 diodes, 1000 transistors, and seventy vacuum tubes. Memory was provided by a magnetic drum that was just shy of twelve inches in diameter and another twelve inches in length spinning at 3000 RPM with 200 tracks. The drum provided 400,000 bits of main memory, and 1800 bits of register memory. CPU registers were provided by using the surface of the drum as a series of delay lines.

MARS-1 was used to figure out how the automation of the rail system would work, and MARS 101 took over the job in February of 1964. MARS 101 handled eight days of reservations, 127000 sets on 238 trains, their schedules, fares, and other important data. The MARS 101s across Japan were upgraded over time, and replacements came in the 102, 103, 104, 105, 201, 202, 301, 305, and 501 (this last entered service in 2002). The 103, however, was significantly different from its predecessors.

Hitachi had been working on transistor technology for some time, and the company’s first fully transistorized computer was completed in April of 1959 and delivered to the Japan Electronic Industry Development Association. This was the HITAC 301 and it was similar to the ETL Mark IV (junction type transistors, synchronous clock at 180KHz, magnetic drum of 1000 words) but it had a larger drum of 1960 words that rotated at 12000 RPM. The HITAC 301 achieved 3.4ms per addition, 4.8ms per multiplication, and 6.4ms per division. Importantly, where the HIPAC was intended for scientific and engineering purposes, the HITAC was intended for business processing work.

In 1949, Japan’s wartime asset freeze, which had prevented monetary flight throughout the Second World War, was lifted. At the same time, the Foreign Exchange Law came into effect with the goal of balancing international payments and stabilizing the country’s currency. A side effect of this was that IBM of Japan couldn’t make remittances to its corporate parent or to the IBM WTC. Some attempts were made both by IBM of Japan and by IBM WTC to ameliorate the situation but these efforts came to no avail. The solution finally came in Q4 of 1960 when the Japanese Ministry of International Trade and Industry accepted a proposal by IBM to license IBM’s patents to Japanese computer manufacturers and in return allow remittances to be made by IBM’s subsidiaries. Hitachi was among the very first Japanese companies to take advantage of this new arrangement along NEC, Fuji Tsushinki Seizo (Fujitsu), Oki Electric, Tokyo Shibaura Denki (Toshiba), Mitsubishi Electric, Matsushita Electric Industrial (Panasonic), and Hokushin Electric Works. Later licensees included Yokogawa Electric, Shimadzu, Shiba Denki, Sharp, Yamura Shinko, and Tokyo Denki Onkyo.

Despite Hitachi having been eager to get such an agreement, they didn’t immediately make any use of it. The next several computers used Hitachi’s own proprietary technologies, methods, and often enough materials (lest we forget that Hitachi was originally a mining technology company, then a construction equipment company, electrical generation, and everything else too).

Of course, the 301 was really only affordable for large enterprises, and this left a market niche unaddressed. That niche was filled by the HITAC 201 which was completed in March of 1961. The 201’s development differed from most other computers at the time by having the goals set largely by the potential customers. Hitachi asked what was most important, and Hitachi’s customers responded with inexpensive magnetic tape drives, a kana-character line printer, at least 4000 word memory, and an instruction set suited for accounting systems. These goals were achieved. The HITAC 201 would ship with Hitachi’s second tape drive system (the first having been developed for the KDC-1 at Kyoto University), magnetic drum memory of 4000 words, paper tape reader, tape-punch, and a line printer capable of 120 lines/minute. The 201 could actually support up to four tape decks with each capable of 1000 digits per second with 300,000 digits per reel. As for computation, this was a fixed-point machine that used base ten, had thirty seven instructions, could perform additions in 4ms and mult/div in 30ms.

The Hitachi HITAC 5020 had begun development in April of 1960 by Kenro Murata and Kisaburo Nakazawa, and this would be a largely diode-based machine operating at about 18MHz. The first prototype was completed in May of 1963, and the first delivery was completed in 1965 to Kyoto University. It used core memory, delay line registers, and swappable cards made up logic elements (not too conceptually dissimilar from the cards made at DEC around the same time).

This was the first computer made by Hitachi to feature an operating system. This system, while primitive, was certainly a welcome addition. It featured a system monitor handling I/O and providing simple hardware abstractions, and a job monitor that would load and run each job automatically. It also included the MCP, or main control program, that acted as the scheduler and interrupt handler. The software package offered FORTRAN IV (optimized for the 5020), Autocode (enhanced with file controls, error handling, I/O buffering, scheduling, and macros), and a set of software libraries for common maintenance, operations, programming, debugging, and more. I cannot find as much information about this system as I could for SHARE, and if any reader knows more, please let me know.

In late April of 1965, Hitachi announced the HITAC 8000 series of computers. These were based upon the Spectra 70 series machines from RCA, and were therefore ISA compatible with the IBM System/360. Like the RCA machines, the Hitachi computers came with a range of capabilities but were upward compatible such that software written for smaller machines would run perfectly well on a larger machine. The 8200 was the smallest, the 8300 and 8400 made up the mid range, and the 8500 was the largest. These were IC-based machines with standardized connectors, standardized I/O devices, and utilized an 8bit data format. This last part was vital for the Japanese market where kana was needed. This was also where the primary incompatibility with both RCA and IBM was found. While the English-language software could certainly be run on a HITAC and Japanese-language software could be run on a Spectra, it would require that all of the software libraries, printing capabilities, and so on also be present. These machines could run POS (Primary OS), TOS (Tape OS), and TDOS (Tape Disk OS) from RCA at the time of their introduction. Based upon these, however, Hitachi introduced DOS, EDOS, and EDOS-MSO. The latter of which was interesting for supporting batch, remote batch, and online processing at rather large scale and offering storage management and automatic file allocation for all three modes.

Hitachi was active in the minicomputer market starting with the HITAC 10 announced in February of 1969. This machine was mostly 16bit, and implemented entirely in ICs (except for memory). In its most humble configuration it featured 4096 words of core memory, but it could be configured with up to 32,768 words of memory. The HITAC 10 featured hardware multiply and divide, and featured up to sixty four I/O channels. The computer could be ordered as a rack mount unit or desktop, and it could make use of many other pre-existing Hitachi peripherals such as tape readers, memory expansions, card or paper tape punch, and so on. The software side of things was quite interesting because it reveals Hitachi’s thoughts about the computers use cases. The HITAC 10 shipped with a macro assembler, FORTRAN, a floating point arithmetic package, a set of mathematic subroutines, a debugger, and a premade calculator program. This minicomputer was clearly targeted toward scientific and mathematical environments, and not so much toward business.

This machine’s successor shipped in 1973 with the HITAC 10II. Physically, it was half the height of the original, cycle times were reduced, and the machine gained support for BASIC. The target market had clearly and quickly changed. Two more updates were shipped, the 10II/A and the 10II/L which added IC memory and floppy disk support respectively. In 1975, Hitachi introduced the HITAC 20. This model’s CPU had more instructions, allowed for up to 256 peripherals to be attached, supported newer CRT consoles, and made communications controllers a standard feature. This made the machine popular as a network controller, and enabled remote connection to mainframes across Japan. This minicomputer was frequently the machine allowing terminals at bank branches to communicate to an off-site mainframe system, thereby greatly lowering the cost of computing in business settings. In the 1980s, the 10II was replaced by the E-600 series of minis, and the 20 was replaced by the E-800 series of minis. These were LSI implementations of higher performance with E-600 being 16bit with full backward compatibility and the E-800 being 32bit with M series compatibility. Both series were of the standard nineteen inch rack mount form factor.

In 1973, Hitachi adopted a large monkeypod tree in the Moanalua Gardens as a corporate symbol. Hitachi promised annual payments of $20,000 to Damon Estate for exclusive rights to the tree’s image. The funds helped to maintain the gardens. The tree is protected by law by the county of Honolulu to this day. This particular monkeypod tree is known as the Hitachi Tree. More recently, John Philip Damon, president of Kaimana Ventures, purchased the gardens from the Estate of Samual Mills Damon for $5.05 million. A new contract was signed between Kaimana Ventures and Hitachi on the 28th of December in 2006. The new fee was $400,000 per annum and would last ten years. This was renewed in 2016. Hitachi seems to really love trees, and various Hitachi subsidiaries have dedicated themselves to protecting rainforests and planting trees around the world.

In 1970, RCA sold its computing division to Sperry Rand, and this forced Hitachi to once again chart its own route in computing. Funny enough, this wasn’t a problem. Three years prior, the Japanese government had funded a project to develop Japanese-designed and Japanese-manufactured high performance computers. So, in 1970, the company announced the HITAC 8700, and the following year, the 8800. These were available on the market in August of 1972, and they featured 32bit virtual addressing, multiple processors, pipelining, and buffered I/O with a dedicated I/O controller. The 8800 was one of the fastest computers available at the time, and offered up to 8MB of core memory. Both the 8700 and the 8800 were used for a wide variety of workloads including time-sharing systems.

To make these advanced computers easier to use, Hitachi developed OS7. This allowed for expansion via virtual memory, and swapping to a high-speed magnetic drum. Like the earlier 8000 series machines, the 8700 and 8800 could operate in batch, remote batch, and on-line interactive modes, but these machines could do so for multiple users and multiple programs simultaneously thanks to their advanced hardware coupled with OS7. The operating system was also aware of the multiple CPUs in the computer and could make accommodations for hardware failures so long as one CPU was still fully-functional. OS7 shipped with FORTRAN, PL/I, COBOL, BASIC, and an assembler. The standard utility suite for OS7 offered a linker, sort/merge, file and volume managers, debuggers, test data generator, batch job manager, terminal handler, symbolic and program library manager, math library, statistic library, graphic routines, and plotter routines.

Development of the HITAC M series of machines began in the autumn of 1971, and these machines took advantage of the new technologies developed for the 8700/8800 including virtual memory, buffers, and the use of a dedicated I/O processor. These were descendants of the IBM S/370 compatible machines, but more advanced and proprietary to Hitachi. The high-end M-180 was announced in November of 1974. It offered two CPUs, up to 16MB of memory, up to 64K of buffer memory for 16 channels, an advanced ALU, and a dedicated array processor for vector calculations. The M-170 could be ordered with only a single CPU, up to 8MB of memory, and up to 32K of buffer memory for eight channels. The M-160 reduced the main memory to 256K to 4MB, zero buffer memory, and just five channels. The M-150 reduced main memory to just 1MB, and channels were reduced to four.

A quick note for those of you who’ve never dealt with mainframes of this era, a channel is comprised of a small processor (usually called a control unit) on its own bus to the CPU. It operates largely independently of the main system processor and can run a channel program (list of 8‑byte channel command words [CCWs]) that can read memory, write memory, address the I/O device on that channel, and then interrupt the CPU when necessary. This design frees the main CPU from having to handle every single event on a peripheral device whether that’s tape drives, hard disks, floppy disks, card readers, teletypes, terminals, or anything else and the CPU can just be made aware of whatever information is coming from the device or going to the device. So, the channel could fetch its own program from memory, interpret the channel commands, issue I/O requests, transfer data over its bus, and signal the CPU when it either completes a command or needs something from the CPU. As the channel handles the attached device, from the CPU perspective it is just sending things to a channel address rather than the device directly. This prevents the extremely expensive computer from sitting idle while handling some I/O event or waiting for data to return.

The HITAC M series were treated to VOS, or Virtual Operating System. Larger machines used VOS2 and smaller machines used VOS1. VOS2 supported virtual memory of up to 16MB, could processes up to fifteen jobs simultaneously, had more complete abstractions of the underlying hardware, supported backgrounding/foregrounding of processes, had a basic concept of permissions allowing for shared resources when specified, and had emulators for DOS, EDOS, and EDOS-MSO. In addition to those emulators, VOS2 shipped with an assembler translator from the 8000 series to the M series, and a COBOL converter for the same.

Smaller machines released by Hitachi in the early 1970s (and in particular the 8250) had used NDOS. By smaller machines, we mean those whose memory was somewhere in the 32K to 64K range. This enhanced DOS offered three main processing modes: single, continuous, and spool. It also expanded general capabilities in day to day use supporting console displays, interactive job control, and better disk space allocation controls. It continued the support of COBOL, FORTRAN, and PL/I but it also added RPG. The real-time functions of the system were enhanced, and better remote batch facilities were added (where the NDOS machine was sending a job to a larger machine such as the 8800).

The NDOS advancements were hybridized with features of VOS2 and this became VOS1. We see batch, on-line, remote batch, and remote batch terminal (the NDOS system) modes all present. Like NDOS, VOS1 supported COBOL, FORTRAN, RPG, and PL/I. VOS1 was more fault tolerant than NDOS with ability to reconfigure devices and flag bad memory/storage for avoidance. VOS1 was largely compatible with NDOS, but not as compatible with VOS2.

In 1976 in the USA, Itel Corporation began selling what were essentially clones of IBM mainframes 370/148, 370/158-1, and 370/158-3. The basic idea being to deliver systems that were equivalent in performance but at lower cost. This attracted quite a bit of attention. Hitachi and National Semiconductor formed a joint venture, National Advanced Systems, in Mountain View and took over Itel’s IBM clone business in October of 1979. This included the production, service, support contracts, inventory, and other assets and operations of Itel’s IBM-compatible business. They then released four new lines of systems named AS/3000, AS/5000, AS/7000, and AS/9000. All of these systems were compatible with System/360, System/370, 4300, and 303X software, and for the System/370, its features were replicated in hardware.

A key breakthrough at Hitachi was their Twin-Well Hi-CMOS process in 1978. Comparing the Hitachi HM6147 to the Intel 2147 on Intel’s HMOS, the Hitachi unit met the Intel’s performance at 87% less power consumption. The team that achieved this was led by Toshiaki Masuhara. Masuhara stated:

Mr. Yasui and I decided to develop the 4K CMOS SRAM. At that time, Intel was the top manufacturer of the high speed SRAM. And the part name was Intel 2147. We decided to develop CMOS SRAM, which was as fast as the Intel nMOS SRAM 2147 and has much less power and much less die area. Since we were using nMOS type polysilicon-load cell for the SRAM cell, the area was very small. They were contained in a single well. And we were using CMOS peripheral, so the power was much lower. I remember it was the year of 1977, we designed the circuit and we manufactured the device at Musashi works by using three-micron process. And we submitted the paper to ISSCC and that was accepted and we had the first paper in 1978. That was made into a project, actually a product later named 6147.

The work by this team proved to be among the most valuable to Hitachi’s computing efforts for the next ten years, for while most of the industry was using NMOS, they were using CMOS.

In September of 1978, Hitachi released the first home microcomputer of Japan, the MB-6880. This little machine was initially named the Page Master MB-6880, but was renamed the BASIC Master Level 1 to differentiate it from other models in the line. This computer was not actually developed by Hitachi’s computer division, but rather by the TV division as this wasn’t considered a serious computer. If anyone has information on precisely how this computer made it to market, I’d love to hear about it. Until we know for sure, I prefer to imagine that it was a drama fit for film.

The 6880 was built around a Hitachi-made Motorola 6800 at 750kHz, 4K RAM, and 8K ROM (featuring an integer BASIC). The display output was either 32 by 24 characters or 64 by 48 pixels. For mass storage, the computer could make use of a 300bps cassette interface. The very next year, the 6880L2 was released doubling the RAM and ROM and adding floating point.

In 1980, Hitachi released the 6881, and here RAM is expanded to 16K and the price is lowered.

The final 6800 8bit from Hitachi was the MB-6885 or Basic Master Jr of 1981. This was a 16K RAM and 16K ROM unit with floating point capable BASIC. It also had dedicated VRAM allowing a resolution of 256 by 192 pixels, the PSU was integrated into the main computer housing, and had a cassette speed of 1200bps. Other immediately noticeable changes are the printer port and expansion port on the rear of the machine.

The sudden name switch to Basic Master Junior is due to the release of the 6809 based Basic Master Level 3 or MB-6890 in 1980. These machines shipped with Microsoft BASIC, and they had far more “professional” peripherals including floppy drives, color displays, printers, and so on. The 6890 was built around the Motorola 6809 at 1MHz, 32K RAM, 24K ROM, and could support a top resolution of 640 by 200 or 320 by 400 in eight colors. This machine also included six expansion slots. This was followed by the 6891 which increased the ROM size allowing for full kana support. The 6892 followed and increased RAM to 64K.

Of course, the world of microcomputers moves quickly. By 1982, an 8bit CPU was a low-end device. Hitachi then released the MB-16001 with an 8088, 320K RAM (expandable), and MS-DOS. This machine also supported color graphics, full kana, and kanji. As this was coming from Hitachi’s computer division who’d by this time realized that microcomputers were, indeed, real computers, the 16001 had FORTRAN and COBOL available.

This, of course, is Hitachi. While they made computers, they also made pretty much everything else somewhere… so very similar machines were created for OA (office automation) and FA (factory automation). These were 8088 machines with RAM ranging from 256K to 512K with 5.25 inch floppy disk drives and 10MB hard disks. They were made available in 1983 as the B-16 range. The main feature to set these apart from anything else available was the software selection. I haven’t read of another computer sold with robotics routines and robot monitoring as standard.

With many different machines capable of taking the high-end market, the Hitachi HD68B09E (Motorola 68B09) at 2MHz was used to create the MB-S1 series of computers in 1984. The S1/10 shipped with 48K RAM, 48K VRAM, and Microsoft BASIC. The form factor was now a desktop machine rather than the wedge, and the formerly expensive and high-end peripherals were quotidian. The S1/20 was essentially the same machine but added a ROM card containing kanji. The S1/30 and S1/40 were the next SKUs in this line, and these integrated the floppy disk drives into the computer’s housing, and these were higher capacity 1MB floppies. In 1985, the S1/10 was released as an S1/10AV adding six voice audio, and Atari-compatible joystick ports. The S1/15, and S1/45 took the 10 and 40 and added a communications ROM and an RS232C port.

While Hitachi had been releasing small machines, the company hadn’t lost sight of the truly large machines either. In August of 1982, the company announced the second Japanese super computer, the HITAC S-810. This computer implemented the IBM System/370 ISA on the scalar Hitachi HITAC M-280H CPU at 35.71MHz with a 256K cache. The vector processor registers were 256 elements wide where each element was 64 bits, and the vector processor ran at 71.43MHz. Main memory was implemented in CMOS ICs, and could be up to 256MB. Expansion memory could be up to 3GB. Total performance peaked at 630 MFLOPS. Just five years later, the company would follow this with the S-820 offering performance up to 3 GFLOPS, 1GB of RAM, and 24GB of extended memory. The S-820 was the first supercomputer to have an integrated video output system as a standard feature.

Both the mainframe and minicomputer HITAC brands continued with the L series. The L-470 and L490 were decidedly more mainframe-like while the L-450 could be order either in a mainframe-like form factor or “Desk-Type.” These machines were implemented in VLSI with 20k gates per chip in the logic circuits. Main memory for these machines could be up to 12MB and maximum disk capacities up to 5.4GB. The most amazing aspect of these machines was that they didn’t require any special power or air-conditioning considerations.

In September of 1985, Hitachi entered the UNIX workstation market with the Creative Workstation 2050 running their in-house HI-UX OS based on 4.2BSD. I’ve never used HI-UX, but it sounds like a great addition to the UNIX ecosystem. The OS added Japanese language support, defaulted to csh, had cross compilers for the 80186/80286 and Motorola 68K CPUs, and shipped with TCP/IP and ethernet support. There’s mention here of a “multi-window screen,” but I do not have much information on which graphics server was in use. In images of the 2050 series, a tiled window manager is clearly seen, but before the end of Hitachi’s workstations, HI-UX was using X11 with CDE. This UNIX lineage still exists at the time of this writing with HI-UX/MPP which makes use of the Mach microkernel. The IBM-compatible mainframe HI-UX releases are of a different lineage entirely coming from Interactive Systems corporation. For the hardware, the 2050 was built around the 68010 with 2MB to 4MB of RAM, 40MB to 65MB HDD, a 3.5 inch floppy disk drive, and either a 15 inch or 20 inch color flicker less, anti-glare treated, adjustable, CRT display.

Simultaneous to the release of the company’s UNIX workstation, Hitachi released a DOS-based workstation, the 2020, based around an Intel 80286, 1MB of RAM, one or two 5.25 inch floppy disk drives, a hard disk ranging from 10MB to 40MB, and a commodity CRT display of either twelve or fifteen inches with the buyer able to select either monochrome or color. For these, Hitachi didn’t ship commodity MS-DOS, but rather they shipped a compatible DOS that offered multi-tasking. I initially thought that this must have been Concurrent DOS. This would explain the presence of cross compilers on the 2050’s HI-UX port as well as DRI provided compatibility libraries for UNIX. This does not appear to have been the case. I have no reason to suspect that Hitachi ever had dealings with DRI, and Hitachi had extensive OS development experience. From everything I’ve been able to find, Hitachi made their own M68K port of UNIX System V release 2 with 4.2BSD, made their own multitasking DOS, and made their own compatibility libraries, cross compilers, and runtimes. If anyone has more information, let me know.

These lines were updated in May of 1988. The 2050/32 brought the 68020 at 20MHz with the 68881 coprocessor and 64K cache, up to 16MB of RAM, and an 88MB HDD. The 2020/32 brought the 80386 with an 80387 and 64K cache, up to 16MB of RAM, and an 80MB HDD. The 2050/32 was released in 1989 with a 68030 at 25MHz, up to 52MB of RAM, and a 168MB HDD. The 2020/32E was released in 1990, and was cost reduced (and physically smaller) with a 386SX at 16MHz, a 387SX, 4MB of RAM, and a 40MB HDD.

With the likes of SGI taking UNIX workstations into the world of graphics, Hitachi chose to compete. This brought about the 2050G in 1988, and the 2050G/ET and 2050G/EX in 1989. The 2050G was a machine built around the 68020 at 20MHz with a 68881 and 64K cache. The machine could be fitted with up to 16MB of RAM, a single 3.5 inch FDD, up to a 2.9GB HDD, and either a fifteen, twenty, or twenty six inch color display capable of 16.7 million colors and a resolution of 1280 by 1024 pixels. The rather high display specifications were enabled by a bespoke LSI chip for two dimensional graphics. The 3D performance was enabled via the same LSI, the 68882, and software. The G/EX used a 68030 at 33MHz with a 68882, increased the maximum RAM to 64MB, and added a RISC coprocessor that functioned as both a 2D and 3D accelerator. The ET was roughly the same as the EX but offered a larger 3.3GB HDD.

The Hitachi Proset 30 was built in 1989 and Hitachi called it a “personal work tool.” It was built around an 8MHz 80286, standard RAM was quite low at just 256K but it could be upgraded to 4.5MB. While the machine wasn’t generous with RAM, it did ship with a 40MB HDD, two 1.2MB FDDs, a 640 by 400 color CRT, SCSI, RS232C, modem, telephone, keyboard, and a mouse. As one might guess by these specifications, the machine was oriented around word processing, drawing, email, phone, and fax. Its OS did allow for window switching, and it came with software for the capture and display of still images.

By 1990, Hitachi had been around for eighty years, they’d entered many industries, and their computing division had developed everything from supercomputers to portables, and their TV division had made some home micros. Notably, Hitachi fabricated their own chips, and they had some good history with fabrication. Their CMOS process was as good as any competitor, and their vertical integration assured them supply of almost all of the components they’d require for any endeavor. The company was well known and well positioned for the competition of the 1990s.

My dear readers, many of you worked at, ran, or even founded the companies I cover here on ARF, and some of you were present at those companies for the time periods I cover. A few of you have been mentioned by name. All corrections to the record are sincerely welcome, and I would love any additional insights, corrections, or feedback. Please feel free to leave a comment.

After the sale of the company, the Bromberg family moved to Los Angeles. Bromberg opened two companies rather quickly. First, he restarted the Irving Bromberg Company, and then in 1934, he started Standard Games. The standout hit this time was Contact. Bromberg’s son Martin went to work for his father after graduating from high school, and he was also drafted into the US Navy for World War 2. His enlistment was inactive as he worked at the Pearl Harbor shipyard in Hawaii while also receiving income from his father’s business. In 1940, Standard Games of Hawaii was opened in Honolulu.

In 1945, Standard Games of Hawaii was sold. Irving Bromberg, his son Martin, and a friend of theirs, James Humpert, then formed Service Games on the 1st of September in 1946. This was a partnership with each of the three providing $50,000 to get things started. The company’s mission was to supply coin-operated games and slot machines to the US military’s facilities across the Pacific. Around this time, Martin Bromberg changed his last name to Bromley.

In 1951, the Transportation of Gambling Devices Act banned slot machines in US territories. The response from Service Games was, on the 15th of February in 1952, to open a distribution office in Tokyo, Japan. The exact arrangement of this venture is murky. There are three distinct names used: Lemaire & Stewart, Japan Service Games, and Service Games of Japan. This business unit, however it was arranged, did quite well. The next year, this unit became Service Games, Japan, Inc and gained its own factories and sales organization.

The first use of the name “Sega” was in 1954 on the Diamond Star slot machine. The company was growing with offices in Panama, and they reincorporated in Nevada. Yet, not all was well. In 1959, the company attracted the interest of both the US Air Force and the Japanese government. Service Games was accused of bribery and tax evasion. While they managed to avoid official charges, the company was banned from US air bases, and the Japanese division was dissolved. The legal battle was resolved in December of 1964 when the US courts ruled that there had been no criminal activity by Bromley or Service Games (testimonies in court are a key source for the early history presented in this article). On the 3rd of June in 1960, Nihon Kikai Seizō, operating as Sega Inc, took over the manufacturing of slot machines, and Nihon Goraku Bussan took over distribution. In 1961, the remaining assets of Service Games were sold for $1.4 million.

David Rosen was born on the 22nd of January in 1930 in Brooklyn. Rosen joined the US Air Force in 1948, and he was stationed in the far East. He was in Shanghai around 1949, but US presence there ended on the 25th of April in 1950. At that point he went to Japan only to be redeployed to Korea shortly after the 25th of June in 1950 with the outbreak of the Korean War. The war ended on the 27th of July in 1953, and Rosen returned to Japan. His service ended in 1954. He started Rosen Enterprises after receiving an honorable discharge, and he was originally focused on exporting Japanese artwork.

Then, he realized that Japan had a serious need for ID photos. A photo ID was needed for school applications, ration cards, railway cards, employment, and almost anything else involving interaction with the government. In the USA, there were automated photo booths, so he began importing those to Japan. Doing this, at that time, involved learning quite a bit in the way of licensing through the Ministry of Industrial Trade and Industry who divided licenses into the importation of necessities, non-necessity desirable goods, and luxuries. Luxury licenses were hard to obtain, but an automated Photo Booth would fall into that second category. It wasn’t really essential but it provided a service for something that was.

With that business moving, Rosen began looking for something to compete for Japanese folks’ disposable income. The most popular entertainments at the time were pachinko, bars, cabarets, and dance halls. Taking inspiration from automated photo booths that he was already selling, Rosen began to look at importing coin-operated games from the USA. There were pinball and air-gun game manufacturers in Chicago, and the market in the USA wasn’t growing much. It was easy to get the machines, but the license from MITI was harder. While it took him a year, he did eventually get the license he needed. He could import $100,000 worth of pinball machines, and due to this limit and import duties of around 200%, Rosen began seeking used machines that he could refurbish. He got the machines, opened a few small arcades, and his establishments were busy from morning to night.

By the 1960s, Rosen Enterprises and Nihon Goraku Bussan had almost the entire market within Japan. In June of 1964, Nihon Goraku Bussan acquired Kikai Seizo, and on the 1st of July in 1965, Nihon Goraku Bussan acquired Rosen Enterprises. The combined company became Sega Enterprises, Ltd. Rosen became chairman and CEO of the new company. I imagine that for NGB, the import licenses that Rosen held were a key factor. He also already had the US relationships. For Rosen, the engineers, factory, and other assets of NGB must have been a major part of his decision.

Sega’s initial focus remained on jukeboxes, gun games, and pinball. Just as had been the case with Rosen Enterprises, these machines were largely second-hand units from the USA that required significant service before being leased or sold. This servicing meant the company needed flippers, bumpers, guns, and really, everything else used in a coin-op. Sega began manufacturing these components and quickly realized that they could simply make their own games.

The first game made by Sega was Periscope. The game had been sketched out by Rosen and was then handed off to Shikanosuke Ochi to assemble a team and develop the game. It was originally a three-player, electromechanical, shooting game. It simulated a submarine attacking warships. The ships were cardboard and hung from chains moving across a backdrop of an ocean. The players would then look through a periscope and fire torpedoes represented by lines of colored lights and accompanied by sound effects. Each player had five torpedoes. At the game’s first demonstration at the 23rd London Amusement Trades Exhibition, Periscope did well and earned $100 which would be about $1000 in 2025 dollars. After some initial sales, customers complained that while the game was extremely popular, it had a poor profit as the machine was expensive at $1295 (a significant portion of that price being a Japanese export tax). The response from Sega was twofold. First, they recommended increasing the price-per-play to 25¢, and second, Sega made a cheaper, smaller, single-player version of the game. Periscope was a massive success, and it largely set the standard of 25¢ per play for arcades in North America while raising the expectations of arcade goers. For Sega, Periscope proved a market and the company made a series of electro-mechanical games of various kinds until 1977.

In 1969, Sega was sold to Gulf and Western. Rosen remained CEO, but many other former owners and executives chose to retire following the sale. In 1970, Rosen became the head of a subdivision of Gulf and Western called Gulf and Western Far East Pacific based in Hong Kong of which Sega was a subsidiary.

Perhaps the most impressive electro-mechanical arcade cabinet that Sega created was Jet Rocket which saw world-wide release in August of 1970. This game didn’t do as well in the USA as it had in Japan due to it having been copied by competitors before it could ship. Jet Rocket was a flight simulator that had neither a proper screen nor any processor. The landscape was on a vertically oriented conveyor belt that scrolled during gameplay. The game surface was then reflected to the player by mirror. Turning the yoke would turn the mirror. When firing, lights in the cabinet reflected orange lights inside the cabinet onto the plexiglass that the player looked through. The reflection on the plexiglass reflected onto the mirror. This provided the illusion of the rockets flying through the air. These rockets were accompanied by sound effects. Hits on the map were illuminated via timing of the map scrolling and fire button presses, but of course, there were no circuits on the map. The illumination of targets and hits were provided by lights behind the canvas, and the timing is what allowed the closure of the circuits controlling the lights. Jet Rocket won several awards, but due to clones there was an oversupply. The game wasn’t the commercial success that Sega had hoped for outside of Japan.

By 1970, Sega’s games were successful enough that the company had committed to opening arcades itself. When the arcades proved to be successful as well, the company entered into a joint venture with Toho Film Studios (maker of Godzilla). The first Sega/Toho location was in Sapporo City, Hokkaido, Japan taking the entire second floor of the Toho Leisure Building. The arcade was divided into two rooms and featured 125 games.

In 1973, Sega released Hockey TV and Pong Tron. Hockey TV was a variation on Pong where the player controlled a vertical line blocker, and the puck had to hit a smaller location on the screen. Pong Tron, on the other hand, was a more traditional Pong implementation. All of Sega’s video game cabinets used discrete logic until 1979.

In 1974, Sega Enterprises Ltd went public in 1974 by becoming a subsidiary of Gulf and Western owned Polly Bergen. Rosen then became the CEO of Polly Bergen which was renamed Sega Enterprises Inc, and Sega Enterprises Ltd in Japan became a subsidiary.

Sega’s first light-gun game was made in 1974, Balloon Gun. Sega described it:

Balloon Gun is a one or two player game and perfect score is 60. With each balloon that is burst by an accurate pistol shot as it appears in the target area, a score of one point is recorded. Each player is given 60 shots and the game ends when that number has been fired. In a two player game, the first one to use all 60 shots terminates the game for both players and the one with the highest score is the winner. As the game starts, the balloons appear in a slow pattern allowing ample time for careful aim but as the game proceeds, the balloons are released at a faster rate adding excitement and making scoring more difficult.

Sound: A cracking sound when the pistol is fired and a “popping” sound when a balloon is hit.

Dimensions: 69 inches by 26 inches by 33 inches

Wattage: 70 watts

In 1975, Sega wanted to replicate their success in Japan in the USA. Kingdom of Oz was an arcade operator in California’s shopping malls, and they’d been reasonably successful but not so much that they’d be too expensive. Sega purchased fifty percent of the company in 1975, and then fully acquired the company in 1976. After acquisition, Kingdom of Oz locations became Sega Centers. After rebranding and redesigning the existing locations, Sega began opening new Sega Centers in 1977.

Gremlin Industries Inc had been founded in 1973, and the company produced the VIC Dual system for video game cabinets, and they also made the Noval 760 microcomputer. The Noval 760 was a commercial failure (despite being pretty cool), but the VIC Dual was good. Sega purchased Gremlin in 1978. Gremlin would eventually become Sega Electronics.

The VIC Dual was comprised of two Zilog Z80 CPUs at 1.9MHz with each capable of 280,000 IPS, 8K RAM, 8 colors from a palette of 64, with a resolution of 256 by 224 at 60Hz. Sound logic was discrete and custom for each game. This board was used between 1977 and 1981 in Safari, Depthcharge, Heiankyo Alien, Car Hunt, Space Attack, Head On, Head On 2, Invinco, Samurai, Deep Scan, Digger, Space Trek, Tranquilizer Gun, Carnival, N-Sub, Pulsar, Star Raker, Borderline, Alpha Fighter, and Special Dual (I, II, III).

The arcade industry was seeing a bit of a lull in the early 1980s, and Sega needed to reduce costs and install times. The answer to these issues was the G80 arcade system. The bulk of the hardware boards could be used by any game, and the specific game board could be swapped. This allowed an operator to change the signage, controls (if necessary), and a single board to change a game. The rest of the cabinet needn’t be modified beyond signage and possibly controls.

The G80 consisted of a card cage and a six slot backplane with around ten different cards available depending upon the game: vector vs raster, a few sound boards, a speech board, two different RAM boards, and the game logic board. As the name implies, this is a Z80 based system. For raster-based systems, this was a Z80 at 8MHz, and for vector-based systems this was a Z80 at 3.86MHz. For sound, the G80 used the Sega SN76496 at either 4MHz or 2MHz, a Sega Melody Generator, and an optional GI SP0250 speech synthesizer. The soundboard also made use of an Intel i8035 MCU at 3.12MHz. Depending upon the raster or vector option a different Sega display controller would be used and these offered 64 to 128 colors on screen from a palette of 256, 28 to 32 sprites with 4 colors per sprite, and a 6bit color depth. Raster systems had 14K RAM, and vector system had 10K RAM.

Distributors didn’t really seem to find the upgradability to be too convincing, and this feature of the G80 didn’t push sales. This was made particularly obvious when the success of Zaxxon, which was completely bespoke, outclassed any of the G80 systems. While Sega had been directionally correct with the idea of an upgradable arcade platform, it was only the JAMMA standards of the middle 1980s that would bring swappable PCBs to arcade systems in a major way.

Sega slightly changed their logo in 1982.

Sega released the System 1 arcade board in early 1983. This board standardized CPU, sound, and graphics. The CPU was a Z80 at 4MHz. The sound system was comprised of Sega SN76496 and Intel 8751. RAM was 20K and ROM was 136K. Graphics for this system well surpassed anything Sega had previously delivered with the Sega 315-5911 sprite line comparator, 315-5012 sprite generator, and 315-5049 tile map generator. The chips working together with the CPU could provide an active resolution of 256 by 224 at 60Hz with a total area of 640 by 260 overscan and 260 scanlines (224 active). System 1 allowed 2048 colors on screen from a palette of 4096 with 16 colors per sprite and up to 32 sprites. This board was used in 4-D Warriors, Block Gal, Brain, Bullfight, Choplifter, Flicky, Garcia, I’m Sorry, Mister Viking, Ninja Princess, Noboranka, Pitfall II, Rafflesia, Regulus, Spatter, Star Jacker, SWAT, Teddy Boy Blues, Up’n Down, Water Match, and Wonder Boy.

In March of 1984, Rosen led a management buyout of Sega with backing from the CSK Corporation. Following the buyout, Isao Okawa (chairman of CSK) became the chairman of Sega, and Hayao Nakayama (former VP of distribution) became the CEO. Jeffrey Rochlis became COO and president. Rosen had announced his intention to step down in November of 1983, but he would continue in a consulting role with the company (and as a shareholder) for quite some time.

Given the down turn in the arcade business that was now worse, Nakayama advocated that Sega enter the home computer and game console business. The Sega SC-3000 microcomputer was released on the 15th of July in 1983. This system was available in red, white, or black and with either rubber keys or a fully mechanical keyboard. The computer was built around the NEC D780C (Z80 clone), 2K RAM, 16K VRAM, TMS9918A video chip, and SN76489A audio chip. The system had no ROM and required a cartridge be present to boot. The computer’s video system could produce 40 columns by 25 rows or 256 by 192 pixels with 32 sprites and 16 colors.

For connectors, the SC-3000 offered RF out, analog out and composite video out via 5pin DIN, RGB out, tape (female 2.5mm), 7pin din printer, a cartridge slot, two joysticks, and power (9V, 850mA). If one opted to buy the Card Catcher, Sega Cards could be used with the SC-3000. The Card Catcher was originally distributed with Zoom 909 and Dragon Wang providing decent value, but the memory cards were limited offering just 4K to 32 KB ROM. The Card Catcher plugged into the cartridge slot.

The 2K RAM limit was a problem, but most cartridges had RAM. If you were wanting to write some BASIC, Sega had you covered with four different options. All of these featured 32K ROM, but contained different amounts of RAM. BASIC Level II A shipped with half a kilobyte of extra RAM. Level II B added 1K, III A added 16K, and III B added 32K.

What is a classic computer without a floppy drive and serial ports? For the SC-3000, these were supplied by the SF-7000 Super Control Station. This expansion unit provided a 3 inch floppy disk drive, an RS-232C port, a Centronics parallel port, 64K RAM, and 8K ROM. It connected to the computer via a short cable that ended with a cartridge port adapter. This essentially required the computer to be placed on top of the expansion unit which I can only imagine made typing a pain (unless the owner also possessed an SK-1100 external keyboard), and it additionally required the expansion to be disconnected to use any cartridge software.

Generally, the SC-3000 was similar to many other micros of the time. It shared CPU, video, and sound with the ColecoVision and Spectravideo SV-318. It shared the same CPU and video as MSX1 computers with the difference between them being the sound chip. MSX computers made use of the GI AY-3-8910 rather than the SN76489A. Yet, these two audio chips were similar enough that Coleco offered a translator in the Coleco SDK. I am unaware of anything like this existing from Sega.

After having started work on their computer, Sega discovered that Nintendo were working on the Famicom. Feeling that they’d need to compete, Sega immediately began working on a cost-reduced version of their system that would do without a keyboard and most expansions while offering a permanently attached joystick. This became the SG-1000. The computer itself was unchanged except that it had just 1K of main RAM, the processor was down clocked to 3.58MHz, and video/audio were output only through RF. It did feature an expansion port on the back of the unit, and the SK-1100 keyboard could be used.

The SG-1000 could make use of all SC-3000 cartridge software except for Music and BASIC (unless the keyboard were attached). It could also make use of the Card Catcher. As the primary joystick was permanently attached, there was only one joystick port for the second player.

On the 31st of July in 1984, just over one year after releasing the SG-1000, Sega introduced the SG-1000 Mark II. This was functionally the same system but with the attached joystick removed and two joystick slots having been added. The default was also a joypad rather than a joystick, and the expansion port was moved to the front of the system.

The Mark I and Mark II had just 68 cartridge games during their lives, 29 Sega Cards, and 26 of the cartridges required the keyboard. The Mark I and Mark II also never saw releases in most countries. They were moderately successful in Japan (even beating expectations and being profitable), and they saw some sales in New Zealand, Australia, Italy, Spain, Singapore, Taiwan, France, and Finland but weren’t the massive success of the Famicom. Partially, Sega had poor timing with both the MSX and Famicom overshadowing the SG-1000, but the company was also completely unfamiliar with the home market. Sega was an arcade company. Thankfully, they learned.

On the 20th of October in 1985, Sega released the Mark III. This system integrated the Sega Card slot, kept the expansion slot on the front, had two joypad ports, and was built around a Zilog Z80A at 3.58MHz, 8K RAM, and 16K VRAM. Audio was provided via the SN76496 PSG and Yamaha YM2413 FM synth. The PSG supported 4 channel mono sound with sample rates from 4kHz to 44.1 kHz, 1bit to 12bit audio depth, and 16 volume levels. The FM chip offered 9 mono FM synthesis channels, and 15 pre-defined instruments. The Mark III utilized the Sega 315-5124 manufactured by Yamaha. From what I’ve been able to find, this was essentially a single-chip version of the System 1 graphics chipset. This chip was 16bit and capable of supporting up to 256 by 240 on screen with an overscan resolution of 342 by 262 at 60Hz with 32 colors on screen from a palette of 64. Sprites could have up to 16 colors with another 16 colors for the background. The Mark III was compatible with the SG-1000 and thus the SC-3000, but I am unaware of any expansion that could transform the Mark III into a full home computer. The system launched at the equivalent of $200 or around $602 in 2025 dollars. Within a year, around a million Marks III had been sold in Japan.

The same year the Mark III launched in Japan, Sega launched Sega of America. With the NES having already launched in the USA, Sega wanted to lower the price of the Mark III for the US market and give it cosmetic changes to appeal to the US audience. This was the “Sega Video Game System” or just “Sega System.” To make the system a little cheaper, the FM synthesis chip was removed, and most units had a game in ROM that would launch if no cartridge was inserted (saving money on bundling a cartridge with the system). Getting ready for CES in the summer of 1986, J. Walter Thompson was hired and given a budget of $9 million for advertising, producing the console packaging, and getting launch titles readied.

When CES came, JWT had an 1800 square feet booth setup for Sega. By all accounts, Sega made a good showing and managed to win a design and engineering award. When the Sega System was released in 1986, it was offered in two different bundles, the Base System which consisted only of the game console a controller for $139, or the Master System which included a second controller and light phaser for $149. As most units sold were of the Master System variety, the name “Sega Master System” became the default term for the international version of the Mark III. This was later made official by Sega.

At this time, Nintendo controlled around 90% of the US home video game market, and Nintendo had distributors like Toys R Us and Sears. After four months on the market, Sega managed to sell around 125,000 units which placed them ahead of the Atari 7800 but well behind Nintendo. By winter of 1987, Sega had released 3D Glasses for the SMS and they’d gained distributors like Macy’s and FAO Schwartz, but Sega was still struggling with sales having reached just 250,000 for 1986 when they’d hoped for around 750,000. Nakayama then chose to invest less money in marketing in North America, and turned over distribution and marketing to Tonka in the USA and to Irwin Toy in Canada.

Most of the Master System’s games library was built of arcade ports which were aided by the SMS sharing much of the Sega System 1 and System 2 platform. This fact was heavily used in marketing. Of course, this was something that Sega was forced to do as well. Nintendo’s developer agreement brought with it exclusivity, and this prevented most Nintendo games from getting ports and prevented many developers from making games for the SMS at all.

In 1987, the international SMS was brought to Japan with a few additions. The Japanese unit brought back the FM synthesizer, had a built in port for the 3D Glasses (rather than using a Sega Card), and included a Rapid Fire unit. The BIOS was also different and with no game inserted would play the Space Harrier theme utilizing both the PSG and FM chip.

The final revision of the Master System from Sega (sort of) was the Master System II. This was a cost reduced system that lacked support for composite output allowing only RF (except in France where the reverse was true), lacked support for Sega Cards and 3D glasses, but actually had a better VDP with more supported modes, improved scaling, and bug fixes. The advantages of the newer VDP are best seen with the Brazilian release of Earth Worm Jim. The SMS II’s built in game was Alex Kidd in Miracle World at launch in 1990, but this was changed to Sonic the Hedgehog in PAL regions after 1992 (NTSC models ceased production that year).

While both the Japanese and North American markets may have disappointed Sega, other markets most certainly didn’t. In South Korea, Europe, and Australia, the SMS was the best-selling console. In New Zealand, it was a close second. In Brazil, well… not only was it the best selling console, it is still being sold today. In Japan, a bit over a million units were sold. In the USA, around 2 million; in Europe, close to 7 million; in South Korea, around 720,000; and in Australia, around 650,000. Then in Brazil, the SMS sold over 8 million.

The Sega Game Gear was released in Japan on the 6th of October in 1990, and it hit the shores of Europe and North American in April the following year. It was, essentially, a handheld SMS retailing for $149.99, and using six AA batteries to achieve around three hours of run time. Sega said five, but from personal experience, I can assure you that that figure was inaccurate. The key selling point against competitors at the time was the color screen and the ability to play actual SMS cartridges through an adapter. The Game Gear achieved just over ten million units sold.

Development of the Mark V began in mid-to-late 1986 and was led by Hideki Sato with Taku Matsubara and Masami Ishikawa contributing heavily along with another five individuals whose names I cannot find. Initially, this design was supposed to be 8bit, building on the SMS, but offloading much of the processing to coprocessors. Then, some information about the Super Famicom started to circulate, and Nakayama stated that something much more powerful was needed, citing the Sega System 16 arcade board built of both the Motorola 68000 (Hitachi FD1094) and Z80 as an example. The Mark V team then worked to replicate as much of the System 16 architecture as was possible. As they worked, they found that the VDP chip design was simply getting to be too large to be successfully fabricated at any kind of reasonable yield, and the design began to be scaled back sacrificing spinning sprites and decreasing the number of colors. As the system was nearing completion, Nakayama chose the name Mega Drive with mega conveying the system’s technological superiority over rivals, and the drive conveying the speed of the CPU. This name wasn’t available for trademark in the USA due to it having been taken by the Mega Drive Systems company who made data storage modules and removable disk drives. In the USA, the Mega Drive became the Genesis.

The Mega Drive was built around the Motorola 68000 clocked at 7.6MHz with 64K RAM, 64K VRAM, 72 bytes of color RAM, and 8K audio RAM. A game cartridge could provide 4MB to 8MB of ROM and an additional 8K to 32K RAM. Sound processing was handled by a Zilog Z80 at 3.5MHz when playing Mega Drive games, but this CPU could also be used for compatibility with SMS games. The Mega Drive featured the YM2612 for FM synthesis and the SN76496 PSG. For the VDP which should now really be called a GPU, the Mega Drive used the Sega 315-5313. This chip was 16bit, could output 60fps at NTSC progressive scan 320 by 224 or 256 by 224; PAL 320 by 224, 256 by 224, 320 by 240, or 256 by 240; NTSC interlaced 320 by 448 or 256 by 448; PAL interlaced 320 by 448, 256 by 448, 320 by 480, or 256 by 480. The GPU could provide 80 sprites on screen with 16 colors per sprite, and it offered integer sprite zooming, sprite priorities, flipping, tiling, and combining. The palette offered was 512 colors at 9bit, but with shadow and highlight this comes out to 1536 total colors. As for colors on screen, the standard was 64. Adding shadow and highlight, the Mega Drive could put 192 colors on screen. If the programmer used a mid-frame palette swap, the Mega Drive could achieve 256 colors. These latter two were only possible utilizing the rather powerful DMA unit on the Sega 315-5313. The DMA could write to VRAM during active display and during blanking, and this allowed the programmer to change colors while drawing a scanline. For a still image, the Mega Drive could output a static bitmap with up to 4096 12bit colors. For the curious, the powerful DMA of the Mega Drive was the source of the “blast processing” hype of Sega’s ads during this era.

The Sega Mega Drive was released in Japan on the 29th of October in 1988 and the entire first production run sold out within two days. Despite strong sales success, the console never rose higher than third place in Japan. The Super Famicom was the most popular, and the NEC PC Engine the second most popular.

Sega felt that they needed a competitor to Nintendo’s Mario, and a competition was held inside the company to create that rival. In November of 1989, Naoto Oshima was working on a game where the player controlled a character who ran through a world of smooth terrains, loops, and jumps. The game emphasized speed, and thus, the original thoughts were around fast animals. Eventually, they settled on a rabbit who could use his ears to grab objects and manipulate his environment. Unfortunately, this idea proved to be too difficult for the Mega Drive hardware. This led to Oshima and his team focusing on animals that could roll themselves into a ball for an attacking move. This narrowed the field of characters to armadillos and hedgehogs. Oshima’s hedgehog idea prevailed. The hedgehog became dark blue to match the Sega logo, but this also helped him standout against lighter blue backgrounds. Oshima then took inspiration from Felix the Cat’s head, Mickey Mouse’s body, and Michael Jackson’s boots from the cover for Bad. That the shoes were red and white comes from Santa Claus. The hedgehog’s names was originally Mr. Hedgehog, but the development team eventually chose Sonic as they felt this name better conveyed a sense of speed (Sonic’s walking speed matched Mario’s running speed), and the team took the nickname Sonic Team. Sonic’s antagonist was a man with a mustache and had originally been one of the possible protagonists. He came in second, and the character was then repurposed as a villain. Mr. Mustache then became Dr. Eggman in Japan and Dr. Robotnik in the rest of the world. The game was programmed by Yuji Naka who’d worked on Phantasy Star and Ghouls ‘n Ghosts. Production began in April of 1990, with Oshima and Naka working around 19 hours per day. Hirokazu Yasuhara came in as a manager and level designer. Four others later joined to help speed development. The game was ready for play testing in early 1991, and it was revealed at CES that year where it won an award for innovation. Sonic was released in the USA on the 11th of June in 1991, in Japan on the 26th of July in 1991, and I cannot find a reliable release date for other regions.

In 1990, Nakayama realized that the North American was quite different from that of Japan, and he wanted to succeed in that market. To that end, he hired an American who’d already proven his capabilities. Thomas Kalinske had successfully revived Barbie and Hot Wheels, and launched Masters of the Universe as CEO of Mattel. He became CEO of Sega of America in 1990. Not being a video games man, Kalinske brought in industry advisors to get him up to speed and adopted the razor and blades model where the console is sold at a low price to push the sales of games. He then created a team to develop games specifically for the North American market, created new advertising campaigns, and switched the bundled game from Altered Beast to Sonic the Hedgehog. All of this was looked down upon by the Japanese board, but Nakayama was insistent that Kalinske be allowed to continue.

Sonic the Hedgehog was released just after Kalinske’s turn around plan, and during the Christmas season of 1991, the Sega Genesis roughly doubled the sales of the Super Nintendo. Over the course of the year, Sega managed to take around 65% of the 16bit console market. This pattern continued for the next four years.

From the start of design, the Mega Drive/Genesis was meant to be expandable. There was the Power Base Converter that allowed SMS games to be played on the Mega Drive, the Modem that allowed users to connect to Sega Meganet (Japan/Brazil, allowed online multiplayer over dialup), the Sega Channel adapter that allowed a user to download games to the adapter’s 4MB of RAM over cable TV, the Mega-CD (or Sega CD) that allowed for games on CD-ROM as well as providing a faster CPU (12.5MHz m68k) and better graphics chip featuring sprite scaling and rotation, and the 32X which provided two 32bit SuperH-2 CPUs at 23MHz with 256K RAM and 256K VRAM. Of all of these add-ons, the Mega-CD was the most popular, but none were incredibly popular. The 32X, in particular, did fairly poorly with 800,000 units sold despite impressive graphical and gameplay capabilities, and despite being able to work alongside the Mega-CD. The full “tower of power” with a Sega Mega Drive, Mega-CD, and 32X was impressive and more fully realized the original vision for the system, but the SuperH-2 CPU suffered production difficulties. Likewise, being difficult to develop for, only six games were available at launch, and the 32X only ever received forty games.

The Mega Drive saw three official variations from its original design. The Mega Drive II is shown above with both the CD and 32X. There was a Mega Drive III that looked quite similar to the later SMS consoles, and of course, there was the Sega Nomad. The Nomad ate up AAs in about the way I drink coffee… quite quickly and in high volume. The stated figure was six AA batteries for four hours of run time. It was, however, a portable Sega Mega Drive which made it quite wonderful. It sold about a million units, and while not a real “success,” it was still a testament to Sega’s capabilities at the time. It had 3.25 inch screen, an AV connector for video out, a mono speaker, headphone jack, and a controller port for a second player. It could not, however, make use of any of the Mega Drive expansions.

The 32X had launched on the 21st of November in 1994, the 3rd of December in Japan, and in January of 1995 in PAL markets. The Sega Saturn launched in Japan on the 22nd of November in 1994. This timing, the lack of availability of the 32X, and this simultaneous release of the Saturn hurt both the Saturn and the 32X.

In 1993, the home video game console market was getting crowded. The SNES, Mega Drive, Atari Jaguar, CD-i, and 3DO were all competing for placement next to home televisions. A team of 27 was assembled and put under the management of Hideki Sato and they began working on the Saturn project. Early rumors were of either a 32bit or 64bit system utilizing CD-ROM, but the only sure thing was that it would likely be capable of arcade-like visuals. The system was decidedly 32bit.

The Saturn was built around two Hitachi SuperH-2 CPUs clocked at 28.63MHz in a master/slave configuration with each CPU capable of 37.22 MIPS and executing two instructions per cycle per CPU. There were then two DMA units, four fixed point math coprocessors, and a 32bit databus. The CD-ROM had its own dedicated Hitachi SuperH-1 CPU as a controller running at 20MHz and achieve 20 MIPS, and this utilized a 16bit databus. Audio was handled by a proprietary sound chip manufactured by Yamaha and had a 16bit databus. This chip was capable of executing four parallel instructions and ran at 22.58MHz. The chip provided the Saturn with 32 audio channels, could handle PCM, FM, MIDI, and LFO, offered both 16bit and 8bit audio depth at a 44.1kHz sample rate, and stereo output. Video was handled by the Sega/Hitachi VDP1 and Sega/Yamaha VDP2 both operating at 28.63MHz. VDP1 handled sprite, texture, and polygon drawing with a 48bit bus width and 16bit word length. VDP2 handled backgrounds, scrolling, and 3D rotation with a 32bit bus width and 32bit word length. Additionally, the Sony CXA1645M was present for RGB composite video encoding. The Saturn had 2MB of “work RAM,” 1.5MB of VRAM, 512K sound RAM, 512K of CD-ROM buffer, and 32K battery backed SRAM. ROM was 512K. The Saturn did have one expansion slot on the top that could add another 512K to 4.5MB of RAM. There were many video modes, but the lowest was 320 by 224 pixels and the highest was 704 by 512 pixels. System RAM achieved 480MB/s, VRAM achieved 286MB/s, and on a 3D rotating playfield, the Saturn achieved a polygon filtrate of roughly a million texture mapped polygons per second. The Saturn supported a color palette of more than 16 million 24bit colors, and the number of colors on screen at one time varied on modes used. Not bad at all for 1994.

Sega sold (at minimum) 170,000 Saturn consoles on launch day in Japan. Pre-orders had sold out and queues were hours long. Sega achieved 500,000 units sold by the end of Christmas in 1994. By June of 1995, Sega had achieved a million Saturns sold. In Japan, Virtua Fighter had been among the most successful arcade games of all time, and that this was the bundled title no doubt helped those sales. Virtua Fighter 2 launched in 1995, and Virtua Fighter 3 followed in 1997. The Saturn became the market leader for a brief time in Japan, and it was the company’s most successful console in Japan. Around 80% of all games produced for the system were Japanese exclusives.

In North America, Sega had felt that the Saturn was too expensive at $399, and they therefore focused on the Sega Genesis add-ons. These, however, weren’t to have a long life, and customers knew this with the launch of the Saturn. This all changed with the launch of the Playstation at $299. Sega of America then prioritized the Saturn’s launch, but they didn’t have enough stock. They chose to prioritize Babbages, EB, Software Etc, and Toys R Us while leaving out Kay-Bee and other retailers. The likes of Sears and Wal-Mart didn’t receive consoles until the summer of the following year, and demand greatly exceeded supply. Customers could, however, easily get their hands on Sony’s Playstation. With the launch of the Nintendo 64 on the 29th of September in 1996, the Saturn was truly moribund in North America. By the close of 1996, Sega had sold 1.6 million Saturns in North America while Nintendo had sold 1.7 million N64s in just three months and Sony had sold 3.2 million PlayStations. The only market to see success with Saturn outside of Japan was Brazil where the PlayStation wasn’t sold at all.

Curiously, other than the Sonic Jam anthology of previous games, Sonic R racing game, and Sonic 3D Blast ported from the Genesis, there were no Sonic titles for the Saturn. Why did the Sega icon not show up? First, Sonic Team was focused on NiGHTS into Dreams. Second, both Sonic X-Treme and Sonic Saturn were under development, but never released. Sonic Saturn was a truly 3D Sonic game not too dissimilar from the later Sonic Adventure, Yuji Naka didn’t like it, so it didn’t happen. For Sonic X-Treme, the game was stuck in development, redevelopment, and management shifts. With other 3D platformers seeing release and success, the bar for quality went up. By the time the game was getting close to ready, the company had begun work on their next console.

A successor to the Saturn was being conceptualized before the Saturn had been released, and this only picked up steam following the introduction of the Sony PlayStation and Nintendo 64. Some decisions were easy to make, and others not so much. Sega wanted to continue using the SuperH CPUs, but they also wanted to make their next console far more easy for developers. Initial GPU development was in the hands of Lockheed-Martin who’d worked on the Model 1 and Model 2 arcade systems, but Sega rejected their proposals. The next consideration was 3dfx’s Voodoo as this would be quite familiar to software developers, and Sega managed to buy around 16% of the company in 1997 as a sign of commitment. By November, that had changed. Sega chose NEC’s PowerVR. This resulted in a lawsuit from 3dfx against both Sega and NEC. This lawsuit was eventually settled in 2000 with 3dfx receiving a bit more than $10 million. To keep developers as happy as possible, the system would ship with Microsoft’s Windows CE (first released on the 16th of November in 1996 for SuperH-3, MIPS 3000, MIPS 4000). Windows CE was initially related to Windows 3.11 and Windows 95 while having some connections to WinPad. Still, there were a few major differences. CE could be shipped as a runtime library not unlike earlier Windows 1 and Windows 2, the system could be shipped to hardware builders as partially source available for easier adaptation to unique hardware platforms, and the kernel could be configured and built either as monolithic or hybrid. On top of these differences, the operating system was also dramatically scaled down in comparison to either 3.11 or 95. For gaming, a main attraction was DirectX 6 support.

After it all shook out, the Dreamcast was built around the Hitachi SuperH-4 at 200MHz. This is a 2-way, superscalar, RISC CPU with 8K instruction cache and 16K data cache capable of 360 MIPS. It had a 128bit SIMD vector unit, 64bit floating point unit, 32bit fixed point unit, a DMA controller, and interrupt controller. The system made use of NEC PowerVR CLX2 at 100MHz. System RAM was 16MB and was accompanied by 8MB of VRAM, 2MB of audio RAM, and GD-ROM buffer of 128K. The system ROM was 2MB. The console could accept media in GD-ROM, CD-ROM, CD-DA, PhotoCD, VideoCD, CD Extra, CD+G, and CD+EG. Audio was handled by the bespoke AICA manufactured by Yamaha which was an improved version of the Saturn’s sound chip. It was capable of 64 PCM channels with a resolution of 16bits or 8bits at a sample rate of 44.1kHz. The sound IC was controlled by an ARM7DI at 2.82MHz. All of the various hardware was connected via the System Bus which offered abstracted interfaces to BIOS ROM and GD-ROM, modem and sound, data transfer between controllers and CPU, the CPU itself, main memory and DMA, and the PowerVR chip. In the case of the Dreamcast, the CPU actually handled quite a bit of the graphics and the PowerVR chip mostly handled rasterizing. Video output could be done via RCA cables, RF modulator, S-Video, SCART, or VGA. Output via VGA was at 480p.

For input, the Dreamcast had several different controllers and four controller ports. The official controller was a variation on the 3D controller from the Saturn, but had two dock connectors. In addition to this controller, the Dreamcast had arcade-style joysticks, lightguns, rod-and-reel motion controller, keyboard, mouse, and flightsticks.

The Dreamcast shipped with a 56K modem, but an ethernet adapter was available (these are rare), and the system was expandable via this port. Iomega had planned and announced a Zip drive for the system, but this was never released. A karaoke expansion was made available in Japan and was quite popular. It had its own expansion slot, and thus the modem could still be used. This allowed users to download songs over the internet.

The dock connectors on the controllers could accept memory cards, visual memory units (VMUs), microphones, and rumble packs. Most users would prefer the VMU to be in the top of the two dock connectors making the VMU screen visible in the controller. For most games, this would just display an icon, but in some it could display player character health or other useful information. A rumble pack would then be in the second slot. When playing games that made use of or accepted voice, a microphone would likely take the place of the VMU. An interesting note for VMUs, is that VMUs had their own batteries, and some games would install mini games playable on the VMU itself when not attached to a controller. These would also sometimes allow data trading by connecting two VMUs together.

The Dreamcast was released in Japan on the 27th of November in 1998 and in the USA on the 9th of September in 1999. The launch in Japan wasn’t great as NEC was struggling to produce enough PowerVR chips. Sega had wanted half a million Dreamcasts available at launch, but the number ended up being just 150,000. Of these, 80,000 were allocated to fulfill pre-orders while the rest were allocated to retailers across Japan. Nearly all of these were sold in the first three days. There were four titles available for the Japanese launch: July, Pen Pen Trilcelon, Virtua Fighter 3tb (along with the arcade stick), and Godzilla Generations. As usual for Sega of Japan, Virtua Fighter was the best seller. About one in three users logged into Sega’s online service within the first few months.

In the US, the launch was far better. The advertising had played up the 9/9/99 for $199 launch date and price, and nineteen titles were available: Soul Caliber, Power Stone, Mortal Combat Gold, Sonic Adventure, Expendable, Blue Stinger, Trickstyle, Hydro Thunder, Tokyo Extreme Racer, Flag to Flag, Monaco Grand Prix, TNN Motorsports Hardcore Heat, The House of the Dead 2, AirForce Delta, AeroWings, NFL 2K, NFL Blitz 2000, and Ready 2 Rumble Boxing. The first day, revenues for Sega of America’s Dreamcast launch totaled $97,904,618.09. Around a quarter million Dreamcasts were sold in just twenty four hours, and more than half a million were sold in the following two weeks. More than 1.5 million were sold in the States by the end of the year. In Europe, the Dreamcast was released on the 14th of October. By Christmas, Sega had sold roughly half a million Dreamcasts, but sales in Europe cooled quickly.

Perhaps the most ambitious game made by Sega for the Dreamcast was Shenmue. It was, at the time, the most expensive game ever produced, and it had extreme interactivity (for example, the main character Ryo {and thus the player of Shenmue} could play a full version of Hang-On or Space Harrier in an arcade), freedom of movement, real-time day/night and weather systems, fully voiced non-playable characters, a large world, and an involved story more commonly seen in cinema. It was an RPG that crossed sub-genres by including adventure, puzzle solving, fighting, and racing, but it was driven by its story. The gameplay mechanics were simple enough to make the game approachable, but it would quickly draw the player into the game world; a world where the player could talk to every single NPC, and those NPCs had jobs, lunch breaks, and homes. They had their own names, ages, and hobbies. Toys seen in the game come from other Sega franchises like Daytona USA, NiGHTS into Dreams, Phantasy Star, and Sonic. These items could be collected. The game was released on the 29th of December in 1999, and it was well received in Japan. Having cost Sega between $47 million and $70 million to create, it sold at a loss.

Overall, the Sega Dreamcast sold 9.13 million units before being discontinued on the 31st of March in 2001. The discontinuation followed a six month loss for the period ending in September of 2000 to the tune of $163 million. That March of 2001, Sega posted a net loss of $417 million. Partially, this was due to the Sony PlayStation 2, but it was also due to a broader softening economy. More games were released for the Dreamcast (619 total), many of them quite groundbreaking and wonderful, but the platform was dead.

The arcade market had been declining, and Sega began closing arcade locations. Most of Sega’s oversees locations were closed including Sega World indoor theme parks.

Sega became a platform agnostic game studio with a small arcade side business. Overtime, Sega had come to be owned largely by itself and private shareholders. CSK still owned 22%. This share was purchased by Sammy making Sammy the largest single shareholder of Sega. Hajime Satomi (CEO of Sammy) then became the CEO of the combined company, Sega Sammy. The portion of this entity that was Sega Enterprises is today Sega Corporation (Kabushiki-gaisha Sega), and is a subsidiary of Kabushiki-gaisha Samī via Sega Sammy Holdings (Sega Samī Hōrudingusu kabushikigaisha). Its major divisions/studios are Ryu Ga Gotoku Studio, Sonic Team, Sega Sapporo Studio, Sega Fave, Sega Amusements International, Sega Studios Shanghai, Creative Assembly, Sports Interactive, Hardlight, Atlus, Play Heart, Two Point Studios, and Rovio Entertainment.

There are so many games that I couldn’t list or talk through, and there are many people who didn’t get a mention. This is always the case. I was a Sega fan in my earlier years, and I owned and played the SMS, Genesis, Saturn, Dreamcast, Game Gear, and Nomad. The SMS, Saturn, Game Gear, and Nomad were gifted from friends who’d lost interest in them. The Genesis was a gift from my father to his four children. The Dreamcast was a Christmas gift for me from my father in 1999. All of them were awesome. It rained and snowed rather often in southern Ohio, so my Segas got quite the workout. These game systems didn’t just bring joy and entertainment, they stimulated my mind to consider just how the wonders of screen and sound were made. I was quite disappointed when the Dreamcast was discontinued, but it is the one computing device with which I’ve never been able to part.

To everyone who ever worked with or for Sega, thank you; you built magic!

My dear readers, many of you worked at, ran, or even founded the companies I cover here on ARF, and some of you were present at those companies for the time periods I cover. A few of you have been mentioned by name. All corrections to the record are sincerely welcome, and I would love any additional insights, corrections, or feedback. Please feel free to leave a comment.

The town of Ivrea is quite old and has a rich history, but today, its population has shrunk from around 90,000 in 1970 to just over 20,000 in the 2020s. In the 1400s, Ivrea gained small Jewish community. By the mid 1800s, Salvador Benedetto Olivetti was a successful textile merchant in that community, and he married Elvira Sacerdoti, a banking heiress from Modena. Salvador and Elvira had Samuel David Camillo Olivetti on the 13th of August in 1868. Tragically, Salvador passed away when Camillo was just a year old. Camillo, whether the result of his father’s death or just his wiring, had a rather solitary nature. He was impetuous, rebellious, nonconformist, and prone to rather sudden outbursts. By all accounts, Elvira did well by her children, Camillo and Emma. They were taught multiple languages, versed in culture and politics, and they were taught to be open-minded. Camillo attended Calchi-Taeggi College in Milan (a boarding school) and then went on to the Royal Industrial Museum (Polytechnic University of Turin today).

Royal Industrial was the first school of its kind in Italy, a school of electrical engineering. Among the prominent minds there was Galileo Ferraris who’d independently developed AC power, invented the induction motor, and taught Olivetti. Olivetti graduated in 1891, and he went to work at a manufacturing firm in London that produced tools for electrical measurement. This was a short lived endeavor, and Olivetti returned to Turin to work for Ferraris. In 1893, Olivetti accompanied Ferraris to the Chicago Electricity Congress and the World’s Columbian Exposition. The two then made their way to West Orange Laboratories in Llewellyn Park, New Jersey to visit Tomas Edison. Olivetti had a positive view of Edison, but he noted that Edison and Ferraris had a difficult time conversing. Ferraris didn’t speak much English, and Edison was hard of hearing. Due to these barriers, Olivetti had to translate for Ferraris… rather loudly.

Ferraris left the United States to return to Italy, but Olivetti chose to continue his adventure. He visited Pittsburgh, Albany, Boston, New York, Salt Lake, and San Francisco. It was the Bay Area that occupied the majority of his time in the USA from what I can tell. He served as an assistant electrical engineer at Stanford from November of 1893 to April of 1894, and he was able to conduct various experiments into the usage and application of electricity.

Olivetti viewed both the British and the American industrial and economic environments favorably, and when he returned to Italy, he sought to build. His first enterprise was to represent the Victor bicycle and Williams typewriter companies in Italy. This didn’t last long. Then with his classmates Dino Gatta and Michele Ferrero, he opened a factory that produced electric metering equipment This company and factory relocated to Milan in 1904. While this had some success, it wasn’t fulfilling for Olivetti.

In 1899, Camillo Olivetti met Luisa Revel. Revel was the daughter of a pastor, and she was rather shy. The two were as opposite as one could imagine, but they fell in love and married. They had six children: Elena (1900), Adriano (1901), Massimo (1902), Silvia (1904), Lalla (1907), and Dino (1912). The children were home schooled for much of their primary educations, and Olivetti wanted them to play and enjoy the sun in their youths as much as possible.

On the 29th of October in 1908, Camillo Olivetti registered Ingegneria Camillo Olivetti & Compagnia with the Notary. The initial funding was 350,000 lire (about 2.9 million USD in 2025 dollars) with Olivetti holding a majority of the company and 13 partners having smaller stakes. With the first factory for his new typewriter company being built, Olivetti wanted a home nearby. He purchased the Convento di San Bernardino which then became Casa Olivetti. The setting of a former convent is fitting. Whether a result of his personal and political beliefs, his wife’s influence, divine revelation, or some combination of these elements, Olivetti converted to Unitarianism. I imagine that in particular, the Unitarian faith’s strict monotheism would have been familiar to him having been raised in Judaism, and the Unitarian emphasis on reason, science, and philosophy would have been attractive to him as an engineer.

Over the course of six months in what was formerly the chapel and now his private study, Olivetti developed the prototype of his first typewriter. Of course, Olivetti didn’t intend to be second-rate. He took another trip to the USA to familiarize himself with the best practices for typewriter production (particularly at Remington, after acquisition that company became Remington Rand), and he returned to Ivrea in late 1909 or early 1910 with Brown and Sharpe automatic lathes and milling machines. His first four employees were Valentino Prelle, Giuseppe Trompetto, Pietro Bronzini and Stefano Pretti. At the World’s Fair in 1911 (April through November) in Turin, the M1 was displayed to the public for the first time. Also on display were the machine tools used in the production of the typewriters, and it was somewhat clear that Olivetti was as proud (if not more so) of his production methods as he was his products. The fair’s catalogue stated that Olivetti was the first and only typewriter factory in Italy, and of the M1 it mentioned that the product was first class, patented in several countries, of an original design, producing legible characters with a standard keyboard, a two-color ribbon, a decimal tabulator, back-space, and margin adjustment.

The first large order for Olivetti machines came shortly after the Fair with the Italian Navy ordering 100. This was followed by another sizable order in 1912 from the Italian postal service. The M1 was usually sold for 550 lire and around 6000 were produced. It was a complicated machine with around 3000 parts, all of which were handmade. The completed typewriter weighed around 37 pounds, had 42 keys, 84 signs. Every description of this machine notes that the M1 was exceptional with rapid operation, high quality construction, and great reliability.

As I have been unable to find the exact date on which that price was set, ascertaining the relative price of the M1 today is difficult. If I assume that this price was set in 1911, the total price in 2025 dollars would be something like $3885.32. I cannot imagine many orders at such a price. Yet, this same price of 550 lire would equate to something around $1200 in 1918. One interesting anecdote was that the M1 was about 100 lire more than a Remington. So, this would place the price of the M1 around $1670 in 1912 adjusted to 2025 dollars.

In running his company, Camillo Olivetti was a generous and familiar man. He implemented a shortened work week, humane factory conditions, high wages, nine and a half months of paid maternity leave, child care, family welfare, worker’s housing, recreation facilities, educational services, and profit sharing. He didn’t separate himself from the workers in his factories, and he maintained friendships with many of them. This camaraderie led to many decisions being made as a group rather than being made top-down. As for his view of the company’s production, he embraced a policy of vertical integration. Product design, research, manufacturing, and sales were done in-house. The parts for every product were likewise produced by the company itself. With a demand for extremely high quality and superior engineering, workers would be educated and trained at Olivetti if they lacked adequate experience. All of this made Olivetti an attractive place to work, and the company grew from 20 employees in 1911 to 110 employees by 1913. Those 110 employees could produce 23 to 28 typewriters each week.

The years 1915 through 1918 were rough on the company with Europe at war. The work week was reduced to 30 hours, workers often had to defer wages to keep the company running, and the factory was converted for wartime needs. At the end of the Great War, the Olivetti company took advantage of some of its recent factory modifications to expand into the office supplies business becoming an Italian equivalent of Remington, a major distinction being that Olivetti’s products were intended to be not only of high quality, but things of beauty. From the minds of designer Marcello Nizzoli, calligrapher Giovanni Mardersteig, and artists Luigi Munari, Ettore Sottsass, Luigi Veronesi, and Gianni Pintori came products that were every bit as visually stunning as they were mechanically sound.

The M20 was unveiled at the Brussels Exhibition of 1920. Compared to its predecessor, the M20 was physically smaller, lighter, and had a fixed carriage. It was also incredibly successful. Four years after introduction, the company’s Ivrea factory had 400 workers producing 4000 units per annum. By 1926, this increased to 500 workers and 8000 units per annum, then to 13,000 per annum by the end of 1929. The design of the M20 wasn’t completely static, and it received updates throughout the 1920s.

On the 22nd of January in 1929, Olivetti’s first international expansion was established with the Hispano Olivetti Company.

The Olivetti family were mildly socialist in their political attitudes, and while Camillo may have converted to Unitarianism, his wife and children remained Waldensian Christians in keeping with their upbringing. In 1922, the National Fascist Party had gained control of the Kingdom of Italy. Camillo Olivetti’s son, Adriano, graduated from Polytechnic University of Turin in 1924. Adriano was opposed to fascism, and he aided in the escape of several political prisoners including Filippo Turati, Ferruccio Parri, and Carlo Rosselli during his first year following university while working as a technical assistant in the engineering group at Olivetti. Like his father before him, he went to the USA, but unlike his father, his trip served to get him away from the authorities. That isn’t to say he didn’t study industry while there, because he most certainly did. For the younger Olivetti, the visit was dominated by Henry Ford’s production lines, and Remington’s organizational systems. Returning to Italy and to the company, Adriano was promoted to the head of mechanical design where he was in charge of the product development processes. In early 1929, he was promoted to the position of technical director. In this position, he was the leader of all engineering and production operations. That same year, the Concordat of 1929 made Roman Catholicism the sole religion of Italy. Unsurprisingly, Adriano suddenly became a member of the party, but we know that this was merely a matter of appearances as the factory in Ivrea offered food, false identities, and shelter for fugitives of the regime; activities which continued until May of 1945.

The Olivetti M40 was released in 1930 replacing the M20. The M40 had keys requiring less force to operate and thus allowed a higher typing speed. It maintained the Olivetti reputation for quality, had a QZERTY layout complete with spacebar, two shift keys, one capital lock key, backspace, and two keys allowing for the changing of the ribbon color. The M40 was in production until 1948, saw numerous updates and revisions, and sold quite well. The M40 was engineered by Camillo, and production was managed by Adriano.

Following the success of the M40, Adriano was promoted to the position of general manager in which he handled the management of all day to day activities of the company. That same year, the company released the Modello Portatile 1, or MP1, create by Gino Martinoli, Adriano Olivetti, Riccardo Levi, Aldo Magnelli, and Adriano Magnelli. This was the company’s first portable typewriter, and it weighed 11.46 pounds. While the M1 and M20 were offered in any color you wanted as long as that color was black, the MP1 was far more lively with red, blue, brown, and green as options.

With the company now competing in office products, typewriters, and mobile equipment, Olivetti continued their international expansion with offices across much of Europe and Latin America. Production tripled from 1933 to 1937, and in 1938, Adriano became the president of the company. Naturally, this production increase required more workers, and those workers needed housing. Adriano then undertook urban planning and infrastructure development in Ivrea creating neighborhoods of three and four story flats with green spaces. These were designed by prominent and well-known architects. From the 1930s through the 1960s, the Olivetti family would invest roughly 3 billion lire in worker welfare in the form of housing, child care, schools, professional training centers, and more. With Olivetti’s presence in the town, Ivrea grew from around 15,000 to more than 30,000 by the late 1950s. That number would triple by the mid-1970s.

During World War II, Adriano’s resistance to fascism grew. On a trip to Switzerland in 1942, he met Allen Dulles who was then the Swiss Director of the US Office of Strategic Services. Adriano became agent 660 of the OSS on the 15th of June in 1943. Mussolini was removed from power on the 25th of July in 1943 through the efforts of Dino Grandi with support from King Victor Emmanuel III. Marshal Pietro Badoglio then became the Prime Minister. Adriano was arrested and imprisoned at Regina Coeli in Rome for conspiring with the enemy, but Italy was in chaos. This chaos allowed him to escape, but he was a fugitive from the law who was variously hiding and running for around six months. Camillo passed away in December, and it is unclear if Adriano was able to be present with his family. Finally in February of 1944, Adriano reached the safety of Switzerland where he once again made contact with Dulles, and he operated his counter-regime efforts from there. Olivetti in Ivrea continued to offer assistance to those fleeing persecution and Adriano supplied the allies with intelligence. In the fighting in Italy, resistance to the regime cost the lives of 24 workers at Olivetti’s Ivrea location, but in the end, the regime was defeated. As peace arrived, Olivetti returned to Italy in May of 1945.

Through all of this, Adriano’s views had solidified to be completely against fascism, oligarchic capitalism, and marxism. He sought something different, and this he formulated as the Community Movement. He believed in a federalism comprised of territorial units that were both culturally homogeneous and economically autonomous. He believed this to be the only path toward uniting industrialization, humanitarianism, and participatory democracy. Outside of politics, his spirit had also moved, and Adriano converted to Catholicism in 1949. These changes in his thinking led him to become the mayor of Ivrea and later to hold a seat in the Italian parliament.

After the war, Olivetti expanded into the calculator market with Divisumma 14 in 1947. The desktop electro-mechanical calculator could perform addition, subtraction, multiplication, and division and print the results. It was the first printing electric calculator to be capable of all four functions and the first to include a negative balance function. The Divisumma 14 also offered a few convenience features. The user could input data merely for reference in print and not calculation, a calculate without print option to avoid polluting the print with unnecessary data, and a stop for avoiding infinite loops such as when attempting to divide by zero. It was offered in beige and blue. The industrial design came from Marcello Nizzoli while the mechanical and electrical design came from autodidact Natale Capellaro. Various iterations on this design were then produced with fewer features at lower price points. Other variations were created specifically for currencies. Later revisions on the design would also add a joystick rather than the two sliders for cursor placement. This design was later refined with Divisumma models 22, 24, and 26.

While Adriano had fled to Switzerland, things hadn’t been quite as easy for Enrico Fermi. Having become a professor in 1926 at the age of just 24, he lacked the claims of vital economic importance that had aided the Olivetti family. In 1938, the Italian racial laws made things difficult and dangerous for his wife, Laura, and several of members of his research team. This pushed Fermi to the USA where he arrived on the 2nd of January in 1939 in New York City. As a well known man, he had offers from five universities, and he chose Columbia. Fermi’s first lecture to the US military about nuclear energy was given on the 18th of March in 1939, and by August of 1941, he was working with six tons of uranium and thirty tons of graphite. Ultimately, Fermi ended up at Los Alamos as associate director.

In 1949, Enrico Fermi was visiting Italy, and that visit included a trip to Olivetti’s factory in Ivrea. Adriano and Fermi discussed several topics, but the most important conversation was one in which Fermi urged Adriano to consider building computers as Fermi felt that the machines would be vital to the future.

In 1952, Olivetti opened a research center in New Canaan, Connecticut to observe US developments in computing (the company’s offices in the USA were in New York City). In 1954, Adriano met Mario Tchou Wang Li in New York. Tchou was born in Rome, spoke Italian, Mandarin, and English fluently, and earned his bachelor’s degree in engineering from the Catholic University of America in Washington, and then earned his master’s in nuclear physics from the Polytechnic School of Brooklyn in 1949. At the time of their meeting, Tchou was working as an associate professor at Columbia. To Adriano, Tchou was precisely the kind of man Olivetti needed. Adriano offered him a job, Tchou accepted. In December of 1954, he arrived in Pisa.

With Tchou as the head of the Laboratorio di Ricerche Elettroniche at Olivetti, the site of the company’s efforts had to be selected. The first location was the Physics Department of the University of Pisa in cooperation with the school’s Centro Studi Calcolatrici Elettroniche. The Olivetti team helped the school complete the Calcolatrice Elettronica Pisana in 1957, remarkable as the first entirely Italian electronic computer.

During the building of the university’s computer, the Olivetti team moved to a nearby villa, and Tchou began recruiting the twelve best young minds he could from the school and surrounding area, and one veteran of the industry, the Canadian Martin Friedman who’d worked on magnetic memory for the Ferranti Mark I. By the end of 1955, the research group numbered about 25. This group immediately set out to build the prototype of a commercial machine, the Macchina Zero also known as the Elaboratore Elettronico Aritmetico 9001, or ELEA 9001. This first machine was a vacuum tube computer, and it was only used internally by Olivetti. Similarly, the 9002 came after this aiming to reduce costs and increase reliability.

Upon completion of the 9002, Tchou gathered his senior researchers and told them that this machine simply wouldn’t do. Olivetti would launch a fully transistorized mainframe computer. For this, the company allied itself with Fairchild, and launched its own transistor company, Società Generale Semiconduttori, in 1957. The two companies then codeveloped the planar process for integrated circuit manufacturing. The prototype transistorized ELEA was completed in late 1958. This became the ELEA 9003 which was presented to the President of the Republic Giovanni Gronchi on the 8th of November in 1959. This machine weighed in at about five tons, and it could run 8000 to 10,000 instructions per second. It was built with transistor-diode logic and core memory. The ELEA 9003 wasn’t built with the concept of words (not really anyway). Each memory location could hold a single alphanumeric character with an instruction being eight characters long. The base memory was thus 20,000 memory locations and could be extended to 160,000 or 20K 8bit instructions or about 26,666 6bit characters. With a cycle time of about 10 microseconds, the computer’s speed was approximately 100KHz. Uptime wasn’t great, just shy of 50% per day, typically being available only between the latter part of each morning to the early part of the evening. This improved over time. The 9003 was capable of limited multitasking with three processes being able to be run simultaneously. At the time of introduction, those programs could be written only in machine language. The machine had no dynamic memory allocation so each program was loaded contiguously and always at the same memory location. By convention, the first 3,000 characters were left available for testing programs. Over time, Olivetti made an assembler, Psico, available along with testing software, a monitor program, and tape handling software. For I/O, the 9003 offered card reader/punch, tape, printer, and of course, an Olivetti typewriter.

The cabinets of the 9003 were made to be fully accessible by a human without the need for ladders standing just shy of five feet high, and the wiring was in overhead conduits rather than under the floor. The contents of each cabinet were color coded with strips that indicated power, memory, ALU, control unit, and the like, and each was arranged in three parts, opening like a book.

Forty 9003s were installed, offered via lease between 1959 and 1964. The first 9003 was installed at Marzotto in Valdagno, and the second was installed at Banca Monte dei Paschi di Siena. This second example is the only currently known to be complete and functional, and it was donated to the Enrico Fermi Technical Institute in Bibbiena by the bank.

The ELEA 6001 was released in 1961 as a smaller and more affordable computer though it was still a mainframe of several cabinets. This computer shipped in two different versions, one for the sciences and one for commercial use, designated by the suffix of S or C to the numeric designation. Characters were four bits which led to some complexities when using machine code (various signifier bits preceding or following a character), and the memory configurations available ranged from 10,000 to 100,000 4bit characters of core with 40,000 being the most common. With little memory on hand and with various schemes aimed at efficient use, the 6001 was shipped with FORTRAN for scientific uses, and with Palgo (Algol dialect) for commercial uses. This machine sold between 140 and 170 units depending upon which source one prefers to trust.

Sadly, Adriano passed away rather suddenly in February of 1960 from a cerebral thrombosis while on a train to Switzerland. Tchou died in a car accident on the 9th of November in 1961 on his way from Milan to Ivrea to discuss a new line of computers built from ICs with management. The two had planned to launch their computers in the United States following Olivetti’s 1959 acquisition of the Underwood typewriter company, but after their deaths, this expansion plan was cancelled. Some workers, documentarians, and Adriano’s personal guard alleged that their deaths were perpetrated by CIA (successor to OSS), but this was countered by the Tchou family who stated there wasn’t any evidence of foul play.

Robert Olivetti was the eldest son of Adriano and was born in Turin on the 18th of March in 1928. He was educated in business administration at Bocconi in Milan, and then at Harvard. He joined the company in 1955, and became the director of the electronics division in 1959. He became CEO in 1962.

By 1960, the research group had moved to Milan, and Federico Faggin joined the company in autumn of that year. He was tasked with designing a small, inexpensive, personal computer. This machine used 1000 logic gates made out of germanium transistors, and it used 200 PCBs. I/O was done with a teletype. The computer was completed in 1961.

From 1962 to 1964, Roberto was trying to keep the computer business running. The purchase of Underwood was a financial burden, and the passing of both Adriano and Tchou had left development idle. With pressure from the US government via ambassador Clare Boothe Luce to sell Olivetti’s electric division to GE, Roberto turned to the Italian government for assistance. The Italian government didn’t view computers as being of any national importance, and thus a bail out was rejected. New board members came in with cash and saved the company, but they favored a sale. Aurelio Peccei became CEO in 1964, and the Olivetti electronics division was sold to General Electric. This sale did not include the typewriter or calculator divisions, and Olivetti retained Underwood which became Olivetti-Underwood. Over the next few years, this would transform from liability to asset as it provided Olivetti access to the US market with established distribution networks.

Faggin’s small computer was refined and adapted by Pier Perotto, Giovanni De Sandre, Gastone Garziera, and Giancarlo Toppi. It became a diminutive, simple, programmable, personal, desktop computer. Given the time, the entire system was built of discrete components: transistors, diodes, resistors, capacitors. For memory, the machine utilized 240 bytes of magnetostrictive, metal-wire, acoustic delay lines with a cycle time of 2.2 milliseconds. All of these components were then mounted on phenolic resin (commonly known as Bakelite) cards. Phenolic resin a was cheap, heat-resistant, nonconductive, synthetic plastic first patented in 1907. It was easy to mold and could be produced quite quickly. The major downsides were swelling under extreme humidity, difficulty in recycling, and toxicity. The material ceased being used with arrival of ABS and PVC.

The only issue that remained for the small team was that this was a computer. Computers were strictly going to be the domain of GE and not of Olivetti. Given that this machine barely qualified as a computer, Garziera spent several nights going through all documentation and references to this product changing the description from computer to calculator. Olivetti was therefore able to keep it. Still, GE now owned the building and everything in that building except for the office in use by this team. The result of this close and uncomfortable proximity was that the four painted the windows on their office so that GE staff weren’t able to see their activity.

Depending entirely on how loosely the definition of computer is used, the Programma 101 was either a programmable calculator or the first personal computer upon its release in 1965. Of course, just viewing the keyboard, the device appears to be a calculator rather than a computer, but this was also true of the KIM-1. The real limitation was in memory where a program of any complexity was nearly impossible.

The keyboard of the P101 was 37 keys, a decimal selector wheel of 0 to 15, and three switches for Program Record, Print Program, and Keyboard release. Input was achieved either through the keyboard or through magnetic cards, and output was achieved primarily through the printer. The “display” such as it was consisted only of two lamps. Solid blue indicated that the machine was ready for input, flashing blue indicated that a program was running, and a red lamp indicated an error. The printer was capable of 30 characters per second.

For all this talk of the machine being a computer, I did say that programs of any complexity were nearly impossible. Well, nearly impossible means that a thing actually is possible. A clever programmer could split a program across multiple magnetic cards and feed them sequentially as each part of a program was run. While this would have been slow, it was possible. Given that this sixty pound, typewriter-sized computer cost $3200 in 1965 which would be around $38,211 in 2025 dollars, someone needing a computer without requisite funds for a larger machine would find this inconvenience acceptable. We know this, of course, from sales.

The total initial production run of the Programma 101 was 44,000 units. The public unveiling of this “calculator” was at the World’s Fair in New York City during the Fair’s second season in 1965. The intended star of Olivetti’s showing was the Logos 27, another of their mechanical calculators. The P101 was in a small backroom, and Olivetti’s management hadn’t really thought it’d receive much attention. The presenter of the P101 informed the audience that he’d be calculating the orbit of a satellite, put the card in the machine, and after a few seconds the computer began printing the result. There was quite a bit of excitement. The P101 was moved to the front of the booth. Fair attendees assumed that there must be hidden wires connecting it to a mainframe offsite, and thus the Olivetti representatives began allowing people a closer look. Following the fair, Olivetti sold 40,000 units in the USA alone. Some of these went to NASA where the humble little machines were used for Apollo 11, as David W. Whittle recalls:

By Apollo 11, we had a desktop computer, sort of, kind of, called an Olivetti Programma 101. It was kind of a supercalculator. It was probably a foot and a half square, and about maybe eight inches tall. It would add, subtract, multiply, and divide, but it would remember a sequence of these things, and it would record that sequence on a magnetic card, a magnetic strip that was about a foot long and two inches wide. So you could write a sequence, a programming sequence, and load it in there, and then if you would—the Lunar Module high-gain antenna was not very smart. It didn’t know where Earth was. So you would have to call up and give the astronauts some—we had two knobs, a pitch and yaw knob, but you have to give him some angles to put it at. Then once the antenna found the Earth’s signal, it would track it, and then you didn’t have to worry. But it had to get within a certain range before it would grab it and track it.

We would have to run four separate programs on this Programma 101, and then in between those programs, we’d have to get out our manuals. I don’t know if you know what a CRC [Standard Mathematical Tables and Formulae] Manual is, but we’d have to look up trigonometric functions and input the data, which today your calculator does that. So what was taking us ten or fifteen minutes to do, today I could do on my hand calculator in ten seconds. Then we would read out the angles that we came up with to the crew, and they would dial them in, look at the signal strength, the signal strength there. They’d go to auto track, and then they could track it. It was a lot of detail stuff like that. I don’t remember any, not just in my systems but other systems, anything that was significant.

The company’s fortunes improved, and Peccei left the company in 1967. Roberto Olivetti then returned as CEO until 1971. After 1971, he remained as the VP and chairman. Between 1965 and 1971, the Programma 101 represented around 24% of Olivetti’s global revenues with that single product bringing in at least $140,800,000. Perotto and his team, with the support of Roberto, likely saved Olivetti.

The P101 was followed by the P602 in 1971, and it was the first Olivetti computer to be marketed as a microcomputer. It used the same architecture as the P101, but used DTL ICs instead of discrete components. It retained the use of delay line memory but doubled the size. One major addition was the system ROM which added trigonometric, logarithmic, and exponential functions. A second addition was an interface for magnetic ribbon memory cartridges holding 56, 112, or 224 blocks where a block was equivalent to four lines of a magnetic card. Finally, the P602 added the IPSO interface (Olivetti Standard Peripheral Interface) allowing the connection of tape readers, punches, plotters, typewriters, hard disks, and other peripherals. This was followed by the P652 which increased memory to 4K and could be expanded to 32K. It also included an integrated magnetic card that could store 192K.

The Divisumma 18 was released in 1972 and was the design of Mario Bellini. This calculator lacked a display and printed all results. The main calculator and printer body was 9.75 inches by 4.75 inches by 2 inches, and after attaching the battery pack, that length was increased to 12.1 inches. While this product didn’t bring in tons of revenue, and it wasn’t all too important in the history of computing, its design is so wild that I had to mention it.

Olivetti released the TC 800 terminal in 1974. This used a two-card Olivetti designed processor and ran the Cosmos operating system. It was followed by the TC 1800 in 1977. Perotto didn’t really seem to want to make terminals, so the same CPU and other internals made their way into the P6060 in 1975. The first and most obvious difference is the keyboard, which is finally a full QWERTY keyboard with a 10-key and some extra keys for specific functions. RAM is 48K with 16K available to the user and an access time of 700ns. The ROM loader is made of bipolar LSI circuits and once loaded into RAM occupies the other 32K. Total RAM could be expanded to allow up to 48K of user-available RAM. The printer is now larger, and the system included a plasma display capable of displaying a single line of 32 characters (upper and lower case characters as well as symbols). That line on the screen could be “scrolled” up and down over a buffer of 80 characters. IPSO peripheral compatibility was maintained, but the P6060 included two eight inch floppy disk drives of 256K, and RS242 and IEEE-488 connectors. One additional accessory was a VDU for use with CRT displays. Software was significantly different as this machine shipped with BASIC, and the printer supported all characters and graphics supported by the system itself as well as framing, scaling, offsetting, axis drawing, and labeling. With these two advantages, the P6060 would have been of great use to anyone needing to create reports or other office documents.

In the later part of the 1970s, Olivetti had around 62,000 employees, around $1.8 billion in revenue (around $9 billion in 2025 dollars), but even more debt with the company losing around $10 million each month. This debt was large enough and expensive enough that the brothers Carlo and Franco De Benedetti were able to purchase 14% of the company for just $17 million. This gave Olivetti, at this time, a valuation of just $121.42 million. It is rather difficult for me to imagine the ire that Olivetti’s prior investors must have felt, but yet, the company could continue.

Carlo De Benedetti was born on the 14th of November in 1934 in Turin. He’s of Jewish descent, and like Adriano, fled Italy for Switzerland during the Second World War. After the war, he returned to Italy, earned his bachelor's degree in electrical engineering from the Polytechnic University of Turin, and then went to work in his father’s business. He did well there, grew the company profits, and led the acquisition of another company. He was then CEO of the combined company until 1976 when he became the CEO of CIR Group, a position he held until 1978 when he became the CEO of Olivetti. Throughout his life, when political turmoil struck Italy, he went to Switzerland. Later in his life, he attained Swiss citizenship. Just prior to the Olivetti acquisition, he was living there with his wife and three sons.

De Benedetti sought to completely reorganize the company and change its focus from typewriters to microcomputers. Nearly all of the company’s senior management and nearly all of the company’s board members threatened to leave. De Benedetti then met with each person individually and thoroughly explained his reasoning. He was successful in persuading them, and the mass resignations never occurred. Much of his plan centered on ending the generous employee benefit programs, cutting mechanical typewriters, renegotiating union contracts, and layoffs totaling a little over 22,000 personnel. De Benedetti’s Olivetti also purchased small stakes in various other companies providing access to more markets, lower parts costs, and access to research.

The company’s office supply business was still going strong with electronic typewriters and calculators bringing in revenue. To this, Olivetti added copiers, and the company expanded into cash registers and ATM machines.

In 1979, the work on a new computer began at the Olivetti Advanced Technology Center in Cupertino. This machine was built around the Zilog Z8001 at 4MHz, 128K base RAM expandable to 512K, two 5.25 inch floppy disk drives supporting 320K floppies, and the Hitachi HD46505 CRTC. For expansion, the M20 offered IEEE-488, RS232-C, and two internal expansion slots. The computer was offered with various upgrades: higher capacity FDDs, memory, a 9.2MB HDD (and accompanying controller card), Corvus Omninet LAN, and a while after release an 8086 CPU card. The choice of the 46505 meant that this machine was vaguely compatible with IBM’s MDA, but that was the only piece of this machine to offer any such compatibility until the release of the Alternate Processor Board (APB, the 8086 card mentioned previously). The M20’s display output was 512 by 256 pixels, and all display modes were graphical. The computer was capable of producing eight colors when using two 32K video memory boards, but this dropped to just four colors from a palette of eight when using a single video memory board. The key advantage for Olivetti in this display setup was that it could support text at 80 by 25 or 64 by 16, and it could more easily support the various languages of Europe.

Given that this machine was initially designed around the Z8001, it lacked a native operating system at the time of design. This was solved with the Professional Computer Operating System or PCOS-8000. This was a single-user, single-tasking, command driven operating system not too dissimilar from CP/M or MS-DOS. Its commands, however, were quite different. For example, vf 1: would format the disk in the left-most disk drive while fl would list the contents of a file. The lack of software for PCOS was well understood, and thus, a CP/M emulator was made available and it was often bundled with dBase II, SuperCalc, and other common software products. The M20 was released in 1982, and it sold well. The turn around had started and Olivetti reported positive net income for 1982 with positive earnings per share around 22¢.

De Benedetti wanted Olivetti computers to have much wider distribution than the company could achieve on its own. In particular, he wanted the company to have broad access to the American market. The first step to this was a truly IBM compatible machine; the company created the M24. The machine was built around an 8MHz 8086 with an optional 8087, 128K RAM expandable to 640K, and a CGA compatible video card that was actually better than plain old CGA. The video card supported 320 by 200, 640 by 200, and 640 by 400 with up to sixteen colors, and it supported an additional mode of 512 by 256 with eight colors. That last mode could be paired with a Z8001 card giving the M24 compatibility with the M20. For ports, the machine had RGB out (25pin), parallel, RS232 (25pin), and a proprietary 9pin DSUB keyboard connector. Later models added another 9pin connector on the keyboard for the attachment of a mouse, and added support in BIOS for 3.5 inch FDDs.

The M24 had seven 8bit ISA slots for internal expansion, but four of these came populated with a second adapter allowing for 16bit cards to be added. The expansions were on a separate board from both the video card and motherboard, and this is where things get… weird. The motherboard was underneath the expansion board and video card, physically separated from them, and mounted upside-down. The motherboard and expansion board then had edge connectors that slotted into the video card which was mounted vertically. Various configurations of this machine were launched in 1983. All of them featured a twelve inch monitor and 5.25 inch 360K floppy disk drive, but that monitor could be green or amber monochrome, or it could be color. Each monitor combination could then be paired with an optional 20MB HDD. The highest end configuration with HDD and color monitor would feature a video card capable of sixteen colors at 640 by 400, while the lower end variants all included a video card capable of just two colors at the same resolution. From what I can tell, the 8087 was available for all models. Pricing on the base model from Docutel/Olivetti in Texas started at $2745 while a more well equipped machine started at $3395.

On the software side of things, the M24 initially shipped with MS-DOS 2.1, a diagnostic program, keyboard drivers, utilities, OLIMENU (TSR GUI similar to TopView), and GW BASIC. Versions of Windows 1 and Windows 2 were available for the M24, and it was supported by Windows 3.0. Given the machine’s compatibility prowess, it could also run CP/M-86, UCSD-P, COHERENT, and XENIX.

In 1983, Olivetti released their first laptop. This was largely the same machine as the TRS-80 Model 100, but it had a physical redesign with a tilting screen. Base RAM was 8K and expandable to 32K. While the Model 100 was a success, the M10 was not. The machine was discontinued after two years.

The US Department of Justice and the American Telephone and Telegraph company (AT&T) agreed on a plan to breakup AT&T on the 8th of January in 1982. That plan would be executed on the 1st of January in 1984. Seeing the loss of revenue from phone rentals on the horizon, AT&T needed new revenue sources, and they needed them as quickly as possible. The attitude within AT&T was quite clear with signs in the company’s offices in 1983 reading with variations on:

There are two giant entities at work in our country, and they both have an amazing influence on our daily lives… one has given us radar, sonar, stereo, teletype, the transistor, hearing aids, artificial larynxes, talking movies, and the telephone. The other has given us the Civil War, the Spanish-American War, the First World War, the Second World War, the Korean War, the Vietnam War, double-digit inflation, double-digit unemployment, the Great Depression, the gasoline crisis, and the Watergate fiasco. Guess which one is now trying to tell the other one how to run its business?

AT&T knew that the products built upon the company’s many innovations in technology were generating large sums of money, and now freed of its legal prohibition from entering the computer market, it wanted to capitalize on those innovations. The company simply didn’t have the time to do this on its own. The M24 was quickly successful in Europe, was a great product, and it beat the IBM PC and XT, competing well with the Compaq Deskpro. Olivetti, meanwhile, needed cash and more market access. On the 21st of December in 1983, AT&T announced the purchase of 100 million shares of Olivetti for $4.26 million. This made AT&T the single largest shareholder at 25% while De Benedetti maintained 15%. Olivetti was now able to market AT&T products in Europe, and AT&T could market Olivetti products in the USA.

Because of this deal, most people in the USA wouldn’t know the M24 as an Olivetti. Had an American used an M24, it would likely have been the AT&T PC 6300 (or less commonly the Xerox 6060, or in France as the Persona 1600 by LogAbax). The 6300 saw a revision in October of 1985 as the 6300 Plus featuring a 6MHz 80286, and this machine could run MS-DOS and UNIX System V concurrently via Simultask. The M24 saw a revision in November 1985 as the M24 SP with a 10MHz 8086. By 1986, Olivetti was the market leader in Europe and the third largest PC manufacturer on Earth. Unfortunately for AT&T the 6300 Plus was a market failure. In late 1986, AT&T ceded all production and development of its PC products to Olivetti.

The British computer market had been somewhat impenetrable for outsiders, but one prominent maker was struggling, Acorn. On the 20th of February in 1985, Olivetti purchased nearly half of Acorn Computers giving the company immediate access to the UK and an astounding amount of technology. One of the first fruits of this acquisition was the Prodest PC 128S which was an Italian localized BBC Master Compact.

The Olivetti M28 was introduced in late 1986 as an AT competitor. It shipped with an Intel 80286 clocked at 8MHz, up to 1MB of RAM (expandable to 7MB), a 5.25 inch floppy disk drive supporting 1.2MB disks, a 20MB to 70MB HDD, and MS-DOS 3.2 or XENIX. The M28 continued to use Olivetti’s enhanced CGA card, and it allowed more expansion options than its predecessor. Pricing started at around $3000, and it was offered in North America as the AT&T PC 6310 and in France as the Persona 1800.

One of Olivetti’s largest competitors in Europe was Triumph-Adler (owner of Pertec, MITS, Omnidata). On the 22nd of April in 1986, Olivetti purchased 98.4% of Triumph-Adler and became the largest office supply company in Europe and held more than 50% of the typewriter market.

From 1987 to 1994, Olivetti produced five distinct lines of workstation machines. The 3000 series (1987) utilized the Motorola 68000 (likely through Thomson or SGS-Thomson with whom Olivetti still maintained various partnerships) and later the Edge Computer variants of the 68000, the 4000 series (1992) utilized the Intel i860 or Motorola depending upon the specific model, the 5000 series (1990) utilized the Intel 486 and later the Pentium, the 6000 series (1991) utilized MIPS, and the 7000 series (1994) utilized DEC Alpha. The early systems shipped with X/OS which was, mostly, 4.2BSD. Later models shipped with Xenix, NT, or VMS depending upon the CPU in use.

Iterations of Acorn machines, rebadged Thomson machines, and Olivetti’s own M28 followed until roughly 1994.

The Olivetti Quaderno (PT-XT-20) was released in 1992 with an NEC V30 running at 16MHz, 1MB of RAM, a 20MB 2.5 inch HDD, PCMCIA, and MS-DOS 5. The display was an LCD with four levels of gray and a resolution of 640 by 400. It ran on six AA batteries. While visually awesome, the most incredible aspect was the machine’s size: 8.27 inches by 5.83 inches by 1.26 inches. Despite having diminutive dimensions the machine contained a sound controller with both input and output and had a built in microphone, a fax modem, serial, parallel, video out, PS/2 keyboard/mouse, and some bundled productivity software. Interestingly, Olivetti made this Quaderno capable of acting as an answering machine. The machine won awards for design but was ridiculed for having very little computing power.

To overcome the deficiencies of the prior model, Olivetti quickly released the Quaderno 33 (PT-AT-60). This machine utilized a 20MHz AMD 386SXLV, 4MB of RAM, a 60MB HDD, and a backlit, seven inch LCD capable of sixteen gray levels at 640 by 480. This model further featured Windows 3.1, Microsoft Works, and Lotus Organizer. The AAs were gone and power came from nickel-cadmium cells. Combined with power management features, this little laptop could manage six hours of operation per charge.

In September of 1994, Olivetti launched Olivetti Telemedia chaired by Elserino Piol. This company then formed a joint venture with Bell Atlantic called Omnitel Pronto Italia with Francesco Caio as CEO. They went on to form Italy’s second GSM mobile network, first privately held, launched in December of 1995. The company had 300,000 subscribers within its first six months of operation.

With the arrival of the CD-ROM, the Pentium, great sound, and high resolution displays, the PC was becoming a multimedia powerhouse. Olivetti saw an opportunity to conquer the living room as they’d once conquered the office. The company sought to create a computing appliance that would deliver audio, video, fax, and a telephone answering machine to the buyer. The product was the Olivetti Envision released in 1995. There were two models with either a 486DX4 at 100MHz or a Pentium at 75MHz. Both were equipped with 8MB of RAM, a Trident TGV9470 with 1MB of video RAM capable of 1024 by 768 with 256 colors, a Crystal CS4231 audio controller paired and an Oak Mozart OTI 605 (Adlib compatible but with an IDE CD-ROM controller), a 1.4MB 3.5 inch FDD, 635MB HDD, CD-ROM, an infrared remote, and an 83key keyboard with an integrated trackball operating over infrared. For external connectors, the Envision featured two SCART ports, audio out, MIDI, VGA out, serial, parallel, and RJ11. Internally, the machine had three expansion slots, and a satellite TV decoder card was available. The Envision could operate in three distinct modes. The first was as a true entertainment device with the remote where a user could control volume, play audio or video CDs, and view photos stored on CDs. The second was with the keyboard where the user navigated the Olipilot shell over Windows 95. The third mode was as a standard Windows PC. The Envision was largely a failure. It was expensive (especially with a Pentium), and it had quite a few bugs.

Olivetti had some rather serious issues during the early 1990s. Competition in the PC market, printer market, and copier market were leading to high R&D costs and forcing the company to accumulate debt. In 1991, the Italian economy went into an inflationary recession, and the government announced austerity measures. Olivetti had been profitable for thirteen consecutive years, but posted a pretax loss of $59 million for the first half of the year on the 11th of November in 1991 following four years of declining revenues. De Benedetti stated to the New York Times that he’d “reassume the direct and complete management” of the company. By 1996, the company had laid off roughly 28,000 people, more than half of their 1989 headcount.

In June of 1996, De Benedetti resigned as CEO and Corrado Passera briefly filled the role. He was then replaced by Francesco Caio on the 4th of July. De Benedetti resigned as chairman on the 3rd of September (though he retained 14% of the company through his company CIR). On the 18th of September, Olivetti held an emergency board meeting and fired Caio. The board appointed Roberto Colaninno as replacement. Several former executives, including De Benedetti, were under investigation by Italian authorities for having understated the losses of Olivetti after Renzo Francesconi who’d only been CFO for six weeks resigned in late August and promptly accused the company of financial impropriety. When the dust settled and investigations ended, no one was ever formally charged with any crime.

On the 12th of March in 2003, Marco Tronchetti Provera, who was chairman of both Telecom Italia and Olivetti, announced his intention to merge the two companies, a move apparently engineered by Colaninno. Olivetti took 51% of Telecom Italia’s shares, and the combined entity became the TIM group.

I’ve omitted many computers and the stories of many people. Partially this is due to length, and partially due to sources, language barriers, and the fact that many weren’t too interesting. I apologize to those whose stories were left out, and I’m sorry if I didn’t cover your favorite machine.

Olivetti’s history can be divided into three distinct parts. First, there were typewriters. As the crisis of this first era mounted, the company moved into microcomputers for its second era. With the IBM PC dominating the world, the company moved into its PC era. When this era came into crisis, the company ultimately ended in a merger, and that entity became a telco. Olivetti was an impressive company that achieved greatness. It was cut down by the same forces that affected so many others. The race to the bottom in PC pricing meant that only the most efficient manufacturers remained standing.

My dear readers, many of you worked at, ran, or even founded the companies I cover here on ARF, and some of you were present at those companies for the time periods I cover. A few of you have been mentioned by name. All corrections to the record are sincerely welcome, and I would love any additional insights, corrections, or feedback (especially for this article as I suck at Italian). Please feel free to leave a comment.

Osborne’s early life was spent at an ashram. An ashram is something like a Hindu monastery, and for the Osborne family, home at this time was the ashram of Sri Ramana Maharshi, the Sri Ramana Ashram, near Arunachala. This setting gave his childhood quite a bit of freedom, nurtured his creativity, stimulated his mind, and provided him self-confidence. It also meant that his first language was Tamil, and his English became distinctly Indian.

Adam Osborne went to Presentation Convent School in Kodaikanal until 1950 when the Osbornes returned to England. The transition was difficult for Osborne. He changed his style of dress, worked to shed his accent, and endeavored to become more British. I cannot imagine that this would have been easy for a child accustomed to freedom, but he went to a Catholic boarding school in Warwickshire for a few years and then began attending Leamington College for Boys in 1954. He earned a degree in chemical engineering from the University of Birmingham in 1961, and he then moved to the USA and earned his PhD from the University of Delaware in 1968. It was while working on his PhD that Osborne learned how to write software. The combination made him an attractive hire for Shell Development Corporation in Emeryville, California.

Osborne’s older sister, Katya, described him as curious and said of him “His interests seemed to constantly shift. A fascination with poetry gave way to science. Science gave way to engineering. Adam wanted to understand and master it all.” Others have described him as handsome, charming, intelligent, quick witted, persuasive, ambitious, restless, and arrogant. The adjectives of charming and arrogant do not exactly get on, and I am led to think that it may have been Osborne’s distinct upbringing that led to the perception of arrogance. Having been raised in a place where nearly all questions were treated as valid and where a young child could walk up to an esteemed religious leader and in familiar language address him with any question, one can easily imagine the type of confidence that child would later possess. He was simply never taught that one should shrink before power. This arrogance or assertive confidence may have been the reason for his firing at Shell. Apparently, he had frequent disagreements with his superiors.

Upon leaving Shell, Osborne took up freelance technical writing, and one of his first jobs in this respect was writing the documentation for the new Intel 4004. Another was The Value of Power written under contract for General Automation in 1972 about minicomputers.

Being in California, particularly within driving distance of Menlo Park, Osborne began attending the Homebrew Computer Club. The first meeting was on the 5th of March in 1975 where a MITS Altair 8800 was present for review. After seeing the Altair, Osborne knew that a guide to using the machine was needed, and he wrote An Introduction to Microcomputers to fill that need. Despite the book having been well written and having filled a need, publishers at the time weren’t willing to publish it. The computer market was tiny and consisted almost entirely of enthusiasts. Osborne knew better. He started his own publishing company, Osborne & Associates Incorporated, in Berkeley, California. Not long after, Bruce Van Natta at IMS Associates (IMSAI) wanted a copy of The Value of Power and called up Osborne asking about it. Osborne informed him that he couldn’t get that book, but that he could get the new microcomputer book. IMSAI bought ten thousand copies at $4 each. One year after having seen the Altair, Osborne’s book had sold more than twenty thousand copies and thirteen schools were using it. His name and reputation quickly grew, and he began a column called “From the Fountainhead” in Interface Age.

In May of 1979, Osborne sold his publishing company to McGraw-Hill where it became Osborne/McGraw-Hill. Under the terms of this deal, Osborne was obligated to stay on until May of 1982, but he wasn’t required to be exclusively dedicated to the company. He could and did pursue other ventures. Brandywine Holdings was started in early 1980. The initial working space was shared with the “Community Memory Project” whose goal was to bring computing to the people via networked computer terminals in public locations. The CMP (founded by Lee Felsenstein, Efrem Lipkin, Ken Colstad, Jude Milhon, and Mark Szpakowski) was largely an electronics-enthusiast anarchist group. The anarchist part of that being quite normal for Berkeley, and the electronics part of that being the nerdy bit.

The hobbyist market had set a standard of the Z80 with CP/M, and two of the three major players, Apple and Commodore, had chosen to go their own ways. Despite that, the standard persisted with Tandy, North Star, Cromemco, Vector Graphic, and Morrow. In Osborne’s mind, both Apple and IBM had proven that one needn’t be the best to make sales, one only needed to supply something adequate that had good support and availability, or as he put it: “adequacy is sufficient; everything else is irrelevant.”

What was standing in the way? The wires. The average computer buyer at the time was expected to plug in multiple boxes, make sure everything was powered and properly configured, and then the buyer could use the machine. Osborne wanted to provide something as simple as a toaster; just plug it in and turn it on. To make this a reality, everything would need to be in a single box, and he wanted it in the smallest package possible. He then realized that this machine could be portable. He looked up the maximum size of a carry-on bag at PanAm, and those became the maximum dimensions of his computer.

He now had his requirements: Z80 CPU, 64K RAM, CP/M, two floppy disk drives, keyboard, serial port, parallel port, and five inch screen. With those requirements established, he hired Lee Felsenstein whom he’d met at HCC, to design the machine in July of 1980. Osborne now had to go figure out the software. He purchased an unlimited CP/M license for $55,000. For BASIC, he traded 2875 shares of Brandywine stock for an unlimited license to CBASIC from Gordon Eubanks, and another 2875 shares for an unlimited license to MBASIC from Microsoft. The details of the WordStar license are unknown, but WordStar was included and we know that there were fees that were not per unit sold. VisiCalc was the standard in spreadsheets but VisiCorp wanted a million dollars for a port and license. Osborne didn’t like that much at all, so he contacted Richard Frank at Sorcim. He asked him to write a spreadsheet program with a few extra capabilities, and SuperCalc was born. This cost Brandywine around $20,000 and 4000 shares, and Sorcim retained the rights to sell the program to others if they paid back the development cost. All told, the software and documentation costs per machine would be a little less than $10 per computer.

By the end of 1980 with $100,000 from Osborne and $40,000 from Jack Melchor, a working prototype of the computer existed, the company had moved into a room loaned from Sorcim, and Tom Davidson had been brought in as VP and general manager as Osborne still had work at McGraw-Hill. Melchor, then insisted that the name of the company be changed to Osborne Computer Corporation to capitalize on Osborne’s name recognition, so in early 1981, Osborne Computer Corporation was officially born.

Through January and February of 1981, Osborne raised $900,000 ($3.2 million in 2025) from venture capital and acquaintances, and he assembled the initial board: Adam Osborne, Jack Melchor, Seymour Rubenstein, Les Hogen, Richard Frank, Robert Bily, Tony Bergis. Georgette Psaris came on as the VP of sales and marketing, Mike Iannamico handled documentation, and Annette Truesdell handled customer and dealer relationships. Frank didn’t get along with Felsenstein, and he claimed a conflict of interest as well. He resigned and Ken Oshman took his place in September.

Pressure was on to have some machines ready for the West Coast Computer Faire held on the 3rd through 5th of April of 1981. Meanwhile, that one room office wasn't working for the quickly growing company. OCC moved into a bare building in Hayward, and people were working and making phone calls in between the construction interruptions going on around them. As the seemingly impossible deadline approached, the first machines were finished days before they were to be demonstrated. The final price for an Osborne 1 was $1795 which would be around $6587 in 2025 dollars. This price may seem high, but WordStar and VisiCalc weren’t exactly free. This price represented a rather high value per dollar. The final configuration of this machine was a 24.5 pound luggable (20.5 inches wide, 9 inches high, 13 inches deep) with a Zilog Z80 clocked 4MHz, 64K RAM, 5 inch monochrome CRT capable of 52 columns by 24 rows, parallel port, IEEE-488 modem port, serial port, dual 5.25 inch floppy disk drives, and software. That software included CP/M, SuperCalc, WordStar, MBASIC, and CBASIC. Optional accessories included an external 9 inch monitor, an acoustic coupler, upgraded floppy disk drives, and a battery pack which provided three to five hours of use.

The machine was heavy, it ran hot, and 16bit machines were right around the corner. Yet, this machine included two disk drives, 64K RAM, a monitor, expansion ports, and software. Were one to so configure a TRS-80, the price would have been higher at the time.

At the faire, we find that OCC was flamboyant. The company had a huge plexiglass booth that towered over others. Steve Jobs called and left a message: “tell Adam he’s an asshole,” at which Osborne laughed and used it as a bit of advertising. Initial impressions of the machine were mixed. People laughed at the screen size (typically until they used it for a bit), but they also generally liked the value. One thing was certain, people were talking about the machine.

The first Osborne 1 shipped on the 30th of June in 1981 to Gerald Wright of Digital Deli in Mountain View with serial number 000001. Jerry Pournelle reviewed the Osborne 1 a bit in the April 1982 edition of BYTE, and said of it:

In the middle of the dilemma, Adam Osborne sent me his new Osborne 1. That darned near solved my problem. Osborne’s machine is good. The first models had some faulty characteristics, but Adam is an honorable man and also smart enough not to risk his reputation by sharp practices. They’re planning retrofits to take care of all major difficulties and most minor ones. The worst of these was the shift lock, which was worse than useless. Then, too, with that tiny screen you needed smooth vertical scrolling (it already had good horizontal scrolling) . There have been some other minor annoyances, but as I said, Adam’s been fixing them. The new Osborne 1 computers - out by the time this is published - will incorporate the improvements, including true three-key rollover and a decent shift lock, and various other fixes. Those who have already bought the machines will be able to get them retrofitted absolutely free. One thing I thought would be a pain turned out not to be. That’s the tiny video screen. Adam has sent me his larger video monitor, which you can connect to the Osborne 1 with a cable, but I find I don’t use it. The little screen turns out to be just at the right focal distance when I sit at the console; and for someone like me, who wears bifocal glasses, that’s a real boon. I carried the Osborne lout to CalTech’s Jet Propulsion Laboratories for the Voyager 2 encounter with Saturn. There were over a hundred members of the science press corps packed into JPL’s Von Karman Center (the press facility) . Most had typewriters. One or two had big, cumbersome word processors. At least one was a terminal connected through a network to the parent system in New York. Nobody had anything near as convenient as the Osborne 1, which is quiet and fast responding.

He went on to describe responses from folks at JPL, and most said the screen was small. After using it, however, Pournelle stated that many were ready to go buy one for themselves. The one drawback that Pournelle saw was the disk format. It wasn’t compatible with the TRS-80. He did, however, note that OCC would be switching to the IBM format in later models.

If you’re wondering about the lag between the machine shipping and a solid review showing up in BYTE, this could easily be explained by OCC’s struggle in meeting demand. Production couldn’t keep up until around February of 1982, which coincidentally was about the same time that Mike Healey started selling Osborne 1s in the UK under the OCC subsidiary there. There were multiple business problems during that time as well. For example, Osborne was expanding into the UK, Germany, and Australia, and arranging that financing wasn’t always smooth.

Tom Davidson was apparently quite unhappy at the company. He went to Taiwan to do some manufacturing negotiation, and while there, gave an interview to some Italian journalists. He vented his frustrations but portrayed Osborne and OCC in a rather negative light. The result was that upon his return on the 24th of September in 1982, he was met at his office by a security guard, escorted to Osborne’s office, and promptly fired. Bob Jaunich was hired as his replacement and took the positions of president and CEO in January of 1983. For the first month, he was quiet and didn’t change much at all. After getting the lay of the land, he began changing things rapidly in February. Most of the prior management was fired and experienced industry professionals were brought on board.

For all this shake up, I am not certain how much of it was to the benefit of OCC. The company’s best quarter was that which ended in March of 1983 with the company having beat their projected sales every month.

Unlike most computer manufacturers of this era, OCC didn’t use distributors. The company manufactured their machines and then shipped directly to retailers. While this did decrease costs, it added a bit of logistics complexity for OCC. That doesn’t seem to have hurt them at all as by June of 1982, the company had nearly 3000 employees and had shipped a bit over 50,000 machines with sales hitting $70 million for the first year of the computer’s availability.

One place that the company struggled, seemingly from the start, was in their accounting systems. At one point in 1981, the system was offline and unable to process anything. This situation wasn’t quickly remedied, and Osborne himself was preparing the 1982 budget on an Osborne 1 in SuperCalc. The result of this is that the company didn’t really have much financial control in place, couldn’t forecast well, and thus, halfway through a project a team may be told they were over budget or out of funds rather unexpectedly. This also meant that spending on quality control couldn’t be allocated well which became a problem in early 1982, and it meant that the company’s overall sales figures weren’t as good as they’d planned and budgeted. So, in the quarter ending on the 29th of May in 1982, the company lost $1.3 million on revenues of $10 million. That month, Adam Osborne was finally able to work at the company full time. The company turned things around with August revenues exceeding $21 million. The issue is, cost of sales were high at $15 million, and after other expenses, the company’s profit was just $290,000. In November, revenue was nearly $30 million and profit was just $298,000.

The Osborne Executive (or OCC-2) was shown to the press in early 1983 with a ban on press until April. The machine was officially announced on the 18th of April in 1983 at a price of $2495. It was bundled with CP/M Plus, MBASIC, CBASIC, WordStar, SuperCalc, Personal Pearl database, and UCSD p-System. The hardware in this machine was upgraded over the OCC-1 with the machine having 124K RAM, a beeper, and a 7 inch amber display capable of 80 columns by 24 rows. The machine was heavier at 28 pounds, and the optional external battery could provide only an hour of run time. The machine now required a fan and air filter, and the case was more square. The handle was also a bit smaller than the original.

This computer had some special tricks. For example, it could run some synchronous communications software that emulated an IBM 3271 terminal for use with IBM mainframes. Other terminal emulation was promised (more IBM models and other terminals), but I do not know if it was ever available. Part of what enabled this was the OCC-2 having fifteen different baud rates. At power up, the OCC-2 would first run a self-test, and it could then be configured with the aforementioned baud rates, multiple cursor options, display reversal, keyboard click, 50Hz or 60Hz power, and character sets.

At this point, the company had about 15,000 OCC-1s on hand and ready to be shipped to retailers. The combined discontinuation of bundling Ashton-Tate’s dBase-II and the announcement of the Executive ahead of availability meant that quite a few retailers cancelled their orders for OCC-1s and ordered Executives instead. This was especially bad as word had made it to retailers ahead of announcement. OCC couldn’t fulfill their orders and this isn’t a good situation for a company that had small profits. Adding fuel to this fire was continued competition from Kaypro, and new competition from Compaq and around 50 other companies making portables. Computers could now be had for between $1000 and $3000 from nearly 300 distinct companies which took away OCC’s price advantage. Roughly half of the company’s international workforce were laid off in early May, and 92 staff members in Hayward were laid off on the 13th of May. Another 93 staff at Hayward were laid off on the 16th of June. On the 2nd of August, another 202 people lost their jobs, their manufacturing plant in Monmouth Junction, New Jersey was closed the following day with about 90 workers there now unemployed, followed by 312 more in Hayward on the 9th of September. The company filed for bankruptcy on the 15th of September in 1983. The company’s poor accounting practices then led to law suits from 24 investors on the 22nd of September. The company had around 80 employees, around $40 million in liabilities, and around 600 creditors. I’ve read some reports of Osborne computers being sold at auctions around this time for as little as $300 per unit.

Ronald J. Brown took over as CEO submitted a plan for exiting bankruptcy to California on the 13th of December in 1983 which was approved in the first quarter of 1984. By this point, OCC was constituted by just thirty full-time employees. Existing stocks of both the OCC-1 and OCC-2 were being sold while the company worked to produce new machines. With less staff, these were sold to users’ groups who began to serve as independent buyers just as distributors and dealers would have for other companies. Osborne was open to working with dealers and other companies, but they were starting from a bad position. Interestingly, rather than offering support themselves, OCC sold one-year service contracts from Xerox.

In 1983, at the West Coast Computer Faire, long before the products would have been available, Osborne had stated to attendees that both a PC-compatible upgrade board for the Executive and a PC-compatible Executive II would be available in half a year or so. Seemingly doubling down on the mistakes made with Executive itself earlier that year. Still, the IBM PC had set the standard and played a large role in the failure of the Osborne Computer Corporation. The company worked to create the Osborne PC, and it was announced in 1984 at the West Coast Computer Faire. The machine resembled earlier OCC machines but was blue, weighed 28 pounds, featured an Intel 8088 CPU at 4.77MHz, packed 256K RAM standard, two 5.25 inch floppy disk drives, two serial ports, a parallel port, a seven inch monochrome amber screen, and RGB out. It could be upgraded with an additional 512K RAM and an 8087. As cool as the Executive II or Osborne PC might have been, it was never released.

The last two Osbornes to be released were the Osborne Encore (OCC-3) and the Osborne Vixen (OCC-4). The Encore was essentially a Morrow Pivot (designed and manufactured by Vadem) with the good and the bad that that implied. It used an 8088, 256K RAM, 16K ROM, a ten inch monochrome LCD (no backlight) capable of 80 columns by 16 lines or 480 by 128 pixels. The Encore was an IBM-compatible, but not fully so; without a backlight the screen could be hard to read in rather normal lighting conditions, and the 16 line display made quite a bit of software nearly unusable. The Vixen was far more recognizably an Osborne with a 4MHz Z80, 64K RAM, 4K EPROM, seven inch amber monochrome display capable of 80 by 24, two disk drives, and weighed in at 18 pounds. It packed CP/M 2.2, WordStar, SuperCalc 2, MBASIC, the Osboard graphics/drawing program, Media Master data interchange program, TurnKey automated setup application, and Desolation (game). The Vixen was priced at $1298 without an HDD. An external 10MB hard disk increased the price to $1495 and with an internal hard disk to $2798. These machines were available in 1985. The OCC-3 wasn’t much of a match from something like the Compaq Portable, and the Vixen was an 8bit CP/M machine in the time of 16bit DOS machines. They failed to bring OCC back.

By 1986, the company’s creditors were attempting to find more money, but all efforts in this arena failed. On the 31st of January in 1986, Jerome E. Robertson was appointed as the Osborne Computer Corporation’s liquidation trustee and he oversaw the sale of the company’s inventory and assets. OCC’s creditors committee asked the bankruptcy court to declare OCC in default on the 5th of February in 1986. A few of the international offices of OCC did continue to operate, but these branches largely sold rebadged IBM compatible machines. Osborne Australia did particularly well until filing for bankruptcy on the 25th of June in 1995. What was left of the company was then sold to Gateway.

Quite a bit has been written about the Osborne Effect. The idea being that the primary reason for OCC’s failure was the premature announcement of future products. It is true that OCC would have sold more OCC-1s had Adam Osborne not revealed multiple future products that were better than the OCC-1 in every conceivable way. This may have prolonged the company a bit, but it wouldn’t have been sufficient. The company had poor bookkeeping practices, an incredibly high CAC, and the company’s timing was brutally bad, releasing an 8bit CP/M machine as the world was transitioning to the new IBM PC compatible standard. OCC’s importance to the industry is not found in the company itself but rather in pioneering the commercial portables market, the bundled software market, and the “spectacle” of releases tradition with their flashy booths at WCCF. Thank you to all of the folks who worked at Osborne Computer Corporation for your contributions to this industry.

My dear readers, many of you worked at, ran, or even founded the companies I cover here on ARF, and some of you were present at those companies for the time periods I cover. A few of you have been mentioned by name. All corrections to the record are sincerely welcome, and I would love any additional insights, corrections, or feedback. Please feel free to leave a comment.

The company’s sales, despite the Depression, hit nearly $1 million by the end of 1934. Wilson then moved to buy the Rectigraph Company in 1935. Rectigraph made a photocopier that used Haloid’s paper, and this would allow the company a bit of vertical integration. Great idea, but the company needed funds. Haloid went public in 1936 to finance the acquisition.

On the 8th of March in 1938, Mosher became chairman of the board. Joseph R. Wilson, who’d led the company’s various expansion efforts, became president.

A quick detour

Chester Floyd Carlson was working for Bell Labs in New York City in 1934 as a patent clerk. He was 28, a physicist with a degree from CalTech, making little money, and had a job that didn’t exactly satisfy him. He’d recently clawed his way out of debt, which gave him some drive, but each day was a grind. He was nearsighted, so he was always bent over his desk to read the papers his job required, and therefore frequently felt pain in his back and neck. Further, he was constantly manually copying patent applications onto carbon paper.

Carbon paper is thin and its coated with wax and pigment. This paper is then put between two sheets of ordinary paper to make a copy of whatever is typed on a typewriter. Alternatives were available such as the mimeograph where a stencil was made on a typewriter, the stencil was then clamped to a cylinder, and as the cylinder turned to pull sheets of paper through, ink was fed onto the stencil leaving a copy of the text on each sheet. This was often messy, error prone, and the mimeograph fluid had a foul odor. The next method for copying was to use a rectigraph or photostat. These took photos of a document, then developed them with the usual photo development processes, differing from normal photography in the paper used and the automation employed. Another wet copy method was offset printing which made high quality copies but involved the creation of a paper master that was transferred to a metal plate and essentially stamped onto sheets. All of these were labor and time intensive while being messy and ill-suited to the type of work Carlson was doing.

In late 1934, Carlson was laid off. He took a job with a patent attorney, got married, and then found a job a year later with P. R. Mallory. All of these various pressures, plus a clear market need, pushed Carlson to work at making a clean, fast, reliable, and high quality copier. As he put himself:

I had my job, but I didn’t think I was getting ahead very fast. I was just living from hand to mouth, you might say, and I had just got married. It was kind of a hard struggle. So I thought the possibility of making an invention might kill two birds with one stone: it would be a change to do the world some good and also a chance to do myself some good.

Carlson chose to anger his new bride, Elsa, and used their kitchen as his first laboratory. After a particularly bad accident spilling sulfur on the stove, Elsa made him move his lab elsewhere. His mother in law owned a building and gave him a room above a tavern in Queens. Of course, experimentation costs money and doesn’t make money, so Carlson was taking courses at New York Law School in the evening to become a patent attorney. To speed up his work, especially as Carlson was now suffering arthritis in his spine, he hired Otto Kornei.

On the 22nd of October in 1938 (this article is late, you see), Carlson and Kornei succeeded. They’d put sulfur on a zinc plate and rubbed the sulfur surface with a handkerchief to build an electrostatic charge. They had a glass slide with India ink on it that read “10-22-38 ASTORIA” and they put the slide on the sulfur surface. With a few seconds of exposure, the primary part of the experiment was complete. They then removed the slide, sprinkled lycopodium powder on the surface, and blew on the sulfur. The excess powder gone, they saw the duplication. They’d invented electrophotography.

Carlson proceeded to attempt to license his patent (2,221,776) to over twenty companies including IBM, Remington Rand, RCA, and GE, and all of them said no. It certainly didn’t help that World War II was underway, and anything not directly involved in the war effort was given little attention. His wife, disapproving of his effort with little to show for it, chose to divorce him.

Still, other things were going well. Carlson had earned his law degree and had become head of the patent department at P. R. Mallory. This put him in a good position when Russell Dayton of the Battelle Memorial Institute dropped by in 1944. The two got to talking and Battelle funded his research and development.

Back to Haloid

While the company had gone public and raised money, the company’s employees went on strike. Wilson stepped in to resolve the dispute, but things were rocky until the war. World War II was great for the company, and they made quite a bit selling photo paper to the military for reconnaissance purposes. Then, of course, the war ended. Every company that had done well selling to the military was faced with a ton of supply, little demand, and fierce competition. Mosher wanted to the sell the company, but Wilson didn’t. Wilson was the son of one of the founders, and he had a son of his own working in the company. For him, Haloid was much more than just a paper company. Wilson won.

In 1945, Haloid’s head of research, John Dessauer, found out about electrophotography and brought it to the attention of Joseph Chamberlain Wilson, the son of Joseph R. Wilson. The younger Wilson immediately saw the value in it, and he and Dessauer went on a trip to Columbus, Ohio to see it. In 1946, Joseph R. Wilson’s son, Joseph Chamberlain Wilson, became president of Haloid. On the 2nd of January in 1947, Haloid and Battelle signed a royalty agreement on the electrophotographic process.

While working to commercialize the technology, a professor at Ohio State University suggested a name change of the process to a combination of xeros and graphein, xerography. Yet, making xerography into a product would cost Haloid $75 million between 1947 and 1960. They borrowed, they sold stock, and Wilson ceased taking a salary and took only shares. Several executives mortgaged their homes to continue funding the project. Carlson himself even moved to Rochester to aid in product development, but he never took a salary from Haloid. Instead, in 1955, Haloid bought his patents in exchange for 50,000 shares of Haloid.

Of course, Haloid did release products using the technology; they just weren’t great. The Model A, nicknamed the OxBox, was a manually operated machine requiring 39 steps and around three minutes to make a single copy. The Model A was followed by the Foto-Flo Model C, Xerox Lith-Master, and CopyFlo among others. None of these did well, but the company was fully invested in the technology. They changed their name to Haloid Xerox in 1958.

The process changed over time. The sulfur was replaced with selenium. Corona wire then applied a uniform electrostatic charge to the plate. The ink became a toner made of iron, ammonium chloride, and plastic. Brushes made from the hair of Australian rabbits were rotated in the machine to remove excess toner, and small air nozzles blew sheets of paper through the machine. All of these advancements were brought together in the Xerox Model 914. The product was introduced on the 16th of September in 1959 at the Sherry-Netherland Hotel in NYC. While one of the two units caught fire, the other worked without error. For the first time, a person could have copies made automatically onto plain paper without any mess (provided the copier didn’t burn the building down).

By this point, the company couldn’t afford a large advertising campaign, but they placed some ads in magazines read by business professionals and ran a few television ads. Other moves by Peter McColough’s marketing team were far more clever. The company placed machines in well-traveled public spaces where it was on display, and in addition to sales, they also offered machine rental for smaller organizations. This was a low price for up to 2000 copies, and each copy after was 4¢. They also promised that a machine could be returned within fifteen days. The 650 pound behemoth was wildly successful. The company changed its name to Xerox in 1961, and began trading on the NYSE as XRX the same year. After a lifetime of struggle and poverty, Chester Carlson was worth more than $200 million. A year before he passed, he gave away more than $150 million spread among the NAACP, the Rochester Zen Center, Dai Bosatsu Zendo Kongo-ji, the University of Virginia, the New York Civil Liberties Union, and the Center for the Study of Democratic Institutions.

Xerox’s crossed $1 billion in revenue in 1968, and Charles Peter Philip Paul McColough became the CEO of Xerox. Wilson was still the chairman, and he’d feared the development of computers would render the text-to-paper/image-to-paper business of Xerox would be rendered obsolete. He’d said to McColough previously: “if we’re going to be big ten or twenty years out, we’ve got to be able to handle information in digital form as well as graphic form.” Wilson was successful in impressing the need on McColough, and McColough felt that Xerox should develop the architecture of the Information Age. To this end, in May of the following year, Xerox bought Scientific Data Systems, a maker of mainframe computers, for $918 million in stock. This brought SDS’s founder, Max Palevsky, onto Xerox’s board as well as Arthur Rock. SDS had never exceeded $10 million in a single year, yet Palevsky became Xerox’s single largest shareholder. In July, Jacob E. Goldman (Jack), chief scientist, went to McColough and said “Look, now that we’re in this digital computer business, we better damned well have a research laboratory.” His proposal went over well as the executive staff all wanted Xerox ranked alongside IBM and AT&T who had the quite famous research centers of Yorktown Heights and Bell Labs. What Goldman saw was one step further than Wilson. He realized that software would lead hardware as the driver of innovation, and that Xerox could profit from the ability of a computer to drive a printer. He envisioned a machine that was half computer and half xerographic printer. The Palo Alto Research Center was the second research center at Xerox, and it was three thousand miles away from Rochester. While this distance helped in giving PARC freedom, it also helped in divorcing the center from the needs of Xerox.

PARC began operation on the 1st of July in 1970. Goldman then chose George Pake to be PARC’s first director, and he split PARC into three units: the Systems Science Laboratory (SSL) under Bill Gunning, the Computer Science Laboratory (CSL), and the General Science Laboratory (GSL) under Pake himself. Pake, in turn, chose Robert Taylor, former deputy director of ARPA’s information processing techniques office who’d shepherded the creation of ARPANET, to run the CSL until someone else could be found. Taylor was exactly the sort of man to recruit the best minds, and he did so. The initial projects included SSL developing laser printing, optical memories (eventually leading to CD-ROM), and speech recognition. The CSL was working toward graphics and a computer system for the research center, and the GSL was working on solid-state technologies.

Early in its life, the fate of PARC was precarious. Xerox had missed its profit targets for both November and December of 1970. The executive staff was now mostly comprised of accountants and financial engineers, and they wanted to control spending. The proposition to kill PARC was put forward. John Bardeen, who’d worked with Brattain and Shockley inventing the transistor, had worked as a consultant for Xerox starting in 1951, and he’d joined the board of directors sometime later. His voice carried weight. At the suggestion to kill PARC, he stated: “this is the most promising thing you’ve got. Keep it!” So, PARC lived.

Taylor had the idea of taking Douglas Engelbart’s oNLine System and moving further with it. He’d helped secure funding for ARC, and he still had contacts at SRI. As people at ARC began looking to leave, it was William K. English, better known as Bill English, who was first to join PARC (he later worked at Sun). English then recruited another twelve engineers from ARC. Taylor’s prior work had also acquainted him with the Berkeley Computer Corporation, and the Berkeley 500. He had purchased the machine for ARPA and shipped it to the University of Hawaii allowing the school to join the ARPANET. That machine’s hardware design was largely the work of Chuck Thacker, and its software the work of Butler Lampson. Both men were recruited as were some of the folks who’d worked with them: Peter Deutsch, Ed Fiala, Richard Shoup, Charles Simonyi (later wrote Multi-Tool Word for Xenix, which became Word). He also hired Alan Kay from SRI. While Kay hadn’t worked on NLS, he’d been present at the demo, and regarded it as one of the greatest experiences of his life. His master’s thesis at the University of Utah in June of 1968 was titled “FLEX: A Flexible Extendable Language” which envisioned a hardware and software system designed for each other. The language wouldn’t be too unfamiliar to us today, but about the hardware, he lamented that nothing existed to fulfill his dream. He wanted a computer with “enough power to out-race your senses of sight and hearing, enough capacity to store thousands of pages, poems, letters, recipes, records, drawings, animations, musical scores, and anything else you would like to remember and change.” Conversations with Kay have been described as a ramble through ideaspace. Years later Kay would better articulate what was running through his head:

Computers’ use of symbols, like the use of symbols in language and mathematics, is sufficiently disconnected from the real world to enable them to create splendid nonsense. Although the hardware of the computer is subject to natural laws (electrons can move through the circuits only in certain physically defined ways), the range of simulations the computer can perform is bounded only by the limits of human imagination. In a computer, spacecraft can be made to travel faster than the speed of light, time to travel in reverse.

Yet, at this time, his vision was absolutely, truly, ahead of everyone around him. During his interview for PARC, he was asked what he thought his greatest achievement would be at the center, and he replied a personal computer. This wasn’t a term at the time, so the interviewer asked, what’s that? He grabbed a nearby notebook, held it up and said, “this will be a flat panel display, there’ll be a keyboard here on the bottom, and enough power to store your mail, files, music, artwork, and books; all in a package about this size and weighing a couple of pounds. That’s what I’m talking about.”

The CSL, like the other two groups, had rented chairs, rented desks, a telephone, and not much else. Of course, to do research on computers, they’d need a computer, and the computer they wanted was DEC PDP-10. At this point, Paul Strassman was replacing every IBM and DEC machine at Xerox with an SDS machine. CSL was brave enough to make the request for DEC, but the request was denied. There was a ton of back and forth, many bridges burnt, and quite a bit of mudslinging, but it boiled down to a denied request. Given this situation, the CSL built the MAXC (Multiple Access Xerox Computer [the C is silent, the name is joke about Max Palevsky]) which emulated the PDP-10. Quite a bit of discrete logic was implemented in microcode, and MAXC used dynamic RAMs rather than core. Those RAMs were 1103s from Intel, and PARC ordered a ton of them. MAXC had 25000 Intel 1103s arranged as 96 on each of 256 boards, with spare completed boards on hand. Many of the chips didn’t work, and this led to PARC’s first creation, a memory tester, a later version of which was used by Intel itself to test RAM chips on the production line. It also led to the implementation of error correction for RAM. In the end, their system ran TENEX, and it was faster than the PDP-10 whose software it was designed to run. It was also quite capable at limited amounts of multitasking, had more accurate floating point (though to run Lisp they actually added the errors back in as Interlisp expected them), and was more reliable. MAXC set the record for uninterrupted availability on the ARPANET, which is amazing for a machine designed and built in about eighteen months for $750k.

Two aspects of building MAXC became extremely important for later work at PARC’s CSL. First, it created a culture of building hardware and software in-house. Second, it established relationships with vendors in the area. Then, this early culture led to a rule: never create a system that isn’t engineered for a hundred users. Timesharing computers? Expect a hundred simultaneous users. Programming languages? A hundred programmers should all be able to easily get going. Personal computers? Expect a hundred novices to be able to sit down and easily get going. Everything had to work at scale, and most importantly, everything had to actually work. During this time period, the CSL also adopted a hiring process. All new recruits were subject to serial interviews with the entire staff, and anyone hired was required to win approval by unanimous, if not entirely enthusiastic, consent.

After the completion of MAXC, Taylor had to hire the actual director of the CSL. I can’t imagine that it’s easy to recruit a person to come be your own boss, but Taylor did. The man he chose, and then all of CSL chose, was Jerome I. Elkind from MIT and BBN. Elkind then recruited some of his own people from BBN chief among whom were Lisp programmers Daniel Bobrow and Warren Teitelman.

Xerox had folks constantly looking at the technology landscape and analyzing everything that may, at some point, be a threat to Xerox’s business model. They wanted to know what the future would be. At the same time, they viewed PARC as insolent due to the PDP-10 issue and the attitudes of people at PARC during the dispute. A major force in this was Don Pendery. The view of Pendery and many at Xerox was one of defending against the future, while PARC’s view, as Alan Kay put it, was “the best way to predict the future is to invent it!” The problem is, corporations really don’t work that way. A corporation needs measurable plans, on paper, that can be presented in a boardroom. In answer to Pendery, George Pake sent the PARC Papers for Pendery and Planning Purposes, thereafter referred to as the Pendery Papers, to Xerox corporate offices. This report is among the most accurate descriptions of the future produced. It described the use of dense magnetic tape for archival and described something like VHS tape, tablet computers with a decently high resolution displays (1024 by 1024 pixels) and wireless networking, voice recognition, widespread use of scanners with OCR, optical storage media, and the networked paperless office. While this was sent to Xerox as a report of what was to come, PARC could also have just said it was what they intended to build.

More than anyone else at PARC, or maybe just more outspoken about it, Alan Kay wanted to build something like what Engelbart and Bush had envisioned. For him, this was his tablet, the Dynabook. The issue was, the development of all of the hardware necessary to make the Dynabook a reality would take a considerable amount of time. He needed an interim device that could do most of what he wanted. Knowing this, Lampson asked Kay if he’d like Lampson, Thacker, and Ed McCreight (McCreight had worked with Rudolf Bayer to invent B-trees) to build him such an interim computer. Kay answered in the affirmative, and on the 22nd of November in 1972, Thacker and McCreight began building the interim Dynabook.

While I have said that Kay really wanted this, that’s not the complete truth. Lampson wanted a cheap PDP-10, Thacker wanted something faster than a DG Nova 800, and Kay wanted his Dynabook. This machine was the Xerox Alto, and it was working by the 1st of April in 1973.

The Alto had a 606 by 808 pixel, monochrome, bit-mapped display, a three button mouse, a five-key chord set, 2.5MB removable HDD, 16-bit highly microcoded CPU built from TTL, 128K RAM (really 64K 16bit words) expandable to 512K when using banking, and it was about the size of an under-counter fridge. And on the 1st of April, it simply booted to a screen that read “Alto lives.” The key was, PARC now had machines that could be built cheaply, and all of these cheap machines were combined with a screen and a pointing device. The display, keyset, and mouse were clearly ideas from NLS, but the keyset never became popular, and the mouse was refined. First, the buttons were nicer, and second, it used a ball rather than wheels.

Everyone at PARC’s CSL was aware of Alohanet (packet radio network at University of Hawaii), and around this time, Thacker was thinking about connecting machines with coaxial cable. He made the statement that coaxial cable is nothing but captive ether, and he proceeded to work on it. Robert Metcalfe (later founder of 3com) and David Boggs then worked on it for a year, and ethernet began to connect Altos at PARC. Some have commented that Alohanet wasn’t the most important part, but rather that Boggs having been a ham radio operator was. He was quite familiar with coercing unreliable things to reliably communicate. Either way, the networked paperless office was being built.

I’ve already mentioned that early DRAMs were unreliable, and this network allowed another issue to be solved. When any Alto was idle, it ran a memory check. Upon the failure of any chip, the Alto would send an alert over the network stating which Alto was faulty, which RAM slot held the bad chip, and which chip in particular was failing. You could be sitting at your desk working while completely unaware anything was broken only to have a repairman show up and state that he needed to service your machine.

Gary Starkweather over in the Webster, NY research center was working on laser printers. The VP of the Business Products Group for Advanced Development, George White, called Goldman to have Starkweather transferred to PARC. Pake was thrilled, and Starkweather moved into the Optical Science Lab (originally part of the SSL). The first laser printer was called EARS for Ethernet Alto Research character generator-Scanning laser output terminal. It was built by Starkweather and Ron Rider. In 1973, it printed documents created on an Alto sent to the printer via ethernet.

Of course, when thinking of the Alto, most people think of the GUI. This wasn’t quite a thing on day one. Initially, the Alto was text oriented.

Of course, this didn’t remain the case. By the time people outside of PARC saw an Alto, it was a GUI oriented machine. If one were to use SmallTalk, the interface was superficially similar to the later Plan9 Rio, and the language was object-oriented and made use of message passing.

There were the WYSIWYG text editors Bravo and Gypsy, Laurel and Hardy email clients, Sil vector graphics editor, Markup paint program, Draw graphics editor (lines and splines), the WYSIWYG circuit editor ICARUS, multiplayer networked gaming, the tiled-window Cedar development environment complete with a strongly typed language (called Mesa) and desktop icons and image viewers, and the Alto had a semi-graphical file manager called Neptune.

The Alto was the first machine to give users a truly simple point-and-click, cut-and-paste, WYSIWYG, WIMP (windows, icons, menus, pointer) environment. It was also the first system to bring together networked workstations with email, ethernet, and printing. The main tools we know today that were missing were spreadsheets and databases.

As great as the Alto and its software were, Xerox didn’t bring them to market. PARC presented the Alto, ethernet, laser printers, word processors (even one in Japanese), and so on, but there was no development organization at Xerox dedicated to taking research prototypes to market. When there finally was in 1976, there still wasn’t a formalized method for transferring technology within the company (that would take another year). Then, as mentioned previously, there was the disfunction between PARC and Xerox corporate that likely made working together impossible anyway.

Adele Goldberg was a researcher within the Smalltalk Group and conceived of a portable workstation. It conceptually borrowed from the Dynabook, but it was a far more realizable goal for the 1970s. A workable and usable prototype was completed in 1978, and around ten were built. The machine was designed around an Intel 8086 running at 1MHz, 256K RAM, a 320K floppy disk drive, and a seven inch, 640 by 480 pixel, monochrome display. It had a mouse, keyboard, 300 baud modem, ethernet, two channel DAC, an EIA-422 interface, and an IEEE 488 interface. The machine could be powered by battery or a wall socket and weighed in at 48 pounds.

Lynn Conway had joined PARC in 1973 already having accomplished quite a bit at IBM in the realm of superscalar design. Far more important than the Alto’s GUI was what the system enabled. Using the graphical environment, the network, and the laser printer, Conway was able to build ICARUS with the help of Carver Mead, Doug Fairbairn, Jim Rowson, and Dave Johannsen. Conway conceived of designing systems using large structured and functional blocks rather than individual transistors or gates. The structures were purposefully designed to be put together in ways that would eliminate wiring delays. Effectively, Conway and Mead had invented the first EDA tool the way we think of them today (the very first having been at IBM in 1966). In June of 1977, Conway then began working on a book using the Alto’s desktop publishing software. This work combined the ideas of Mead and Conway, included an ICARUS tutorial written by Fairbairn and Rowson, and included design examples made by Johannsen. The book was titled Introduction to VLSI Systems. The first classes to use portions of the book were held in 1978 by Conway as a visiting professor at MIT, the book was published in 1979, and it became the standard text for education in the field. It was then translated into Japanese, Italian, French, and Russian, and was being used in around 120 universities around the globe. Fairbairn went on to cofound VLSI Technology with his company obviously using the technology. Jim Clark learned the VLSI design system from Conway while spending a summer at Park, and went on to design the Geometry Engine and found Silicon Graphics. Of course, as Conway went off to MIT, the Alto made it out of PARC. Altogether, around fifty Altos were donated to MIT, Stanford, and Carnegie-Mellon. The machines were put to work in the research communities of those schools, often in AI work, and they were certainly used for VLSI design courses. Conway left PARC for DARPA’s Strategic Computing Initiative in 1983.

In the autumn of 1979, Jef Raskin and Bill Atkinson had been working on the Lisa at Apple for about a year, and they had been urging Ken Rothmuller, the project manager, to use a bitmapped screen. The Lisa had working graphics from an early point in its development thanks to Atkinson’s prior work on the Apple II Pascal’s graphics. LisaGraf, which would later become QuickDraw, had been created by the spring of 1979, and a basic user interface existed shortly later. The teams at Apple were well aware of PARC. Atkinson had learned of SmallTalk as an undergrad, Raskin had worked at PARC, a few Apple employees had friends working there, and some people at Apple had worked for Engelbart at SRI. The real problem for these teams was in convincing leadership at the company of the value of these technologies. Unbeknownst to most staff was that earlier that year, Xerox had invested $1 million for 100,000 pre-IPO share of Apple. In exchange for this, Apple wanted access to PARC. Raskin arranged two visits, and Apple’s engineers were treated to all of PARC’s various innovations. These weren’t simple demonstrations, but rather full walkthroughs of the technologies, how they worked, and how they integrated together to build a networked office. Bill Atkinson and Andy Hertzfeld themselves aren’t too sure on the chronology, but Hertzfeld believes that Apple had a mouse-oriented windowing system before the PARC visit. We do know, for sure, that such a system existed by the spring of 1980, and further that all vestiges of a soft-key based user interface were gone by that summer. Steve Jobs was on the second visit in December, and of the visit itself, Steve Jobs later stated:

I had three or four people (at Apple) who kept bugging that I get my rear over to Xerox PARC and see what they are doing. And, so I finally did. I went over there. And they were very kind. They showed me what they are working on. And they showed me really three things. But I was so blinded by the first one that I didn’t even really see the other two. One of the things they showed me was object oriented programming – they showed me that but I didn’t even see that. The other one they showed me was a networked computer system… they had over a hundred Alto computers all networked using email etc., etc., I didn’t even see that. I was so blinded by the first thing they showed me, which was the graphical user interface. I thought it was the best thing I’d ever seen in my life. Now remember it was very flawed. What we saw was incomplete, they’d done a bunch of things wrong. But we didn’t know that at the time but still thought they had the germ of the idea was there and they’d done it very well. And within – you know – ten minutes it was obvious to me that all computers would work like this some day. It was obvious. You could argue about how many years it would take. You could argue about who the winners and losers might be. You could’t argue about the inevitability, it was so obvious.

The goal of all of that urging was accomplished. Apple executives too became believers in a graphically oriented future. They built upon and refined these ideas creating something totally new, quite elegant, and with Macintosh, quite successful. At least one person at PARC knew that giving people such access wasn’t the best idea. Adele Goldberg argued vociferously against allowing Apple in a meeting over the course of three hours. She later stated:

He (Steve Jobs) came back and I almost said asked, but the truth is, demanded that his entire programming team get a demo of the Smalltalk System and the then head of the science centre asked me to give the demo because Steve specifically asked for me to give the demo and I said no way. I had a big argument with these Xerox executives telling them that they were about to give away the kitchen sink, and I said that I would only do it if I were ordered to do it, cause then of course, it would be their responsibility, and that’s what they did.

On the other hand, Larry Tesler was delighted by Apple. He said that the Apple team more well understood the technology and what it meant than anyone among Xerox’s leadership in just an hour. Tesler joined Apple in 1980.

Naturally, the whole matter of the GUI became quite litigious. Early on in the life of the Macintosh, Apple needed software developers. The company gained a friend with Microsoft to whom they licensed many of their GUI technologies. Apple didn’t seem to care when Microsoft released Windows 1.0, but they certainly cared when Microsoft released Windows 2.0. On the 17th of March in 1988, Apple filed suit against Microsoft and Hewlett-Packard in San Francisco over the look and feel of the Macintosh. On the 25th of July in 1989, Judge William Schwarzer stated that of the 189 features Apple claimed Microsoft to have copied, all but ten were covered by the license Apple had granted. Then on the 14th of April in 1992, Judge Vaughn Walker ruled those remaining ten elements uncopyrightable. When this finally ended in 1994, the courts were in favor of Microsoft due to prior art on the overall concepts and due to the license having been granted.

IBM’s first copier had been released in April of 1970, and it wasn’t a good product. It did, however, have the name of IBM behind it. Xerox quickly chose to sue IBM for patent infringement and this was settled in 1978 with IBM having paid Xerox $25 million. Unfortunately, Xerox was then under fire from the FTC, Kodak began competing with them, Canon entered the market, and Ricoh entered the market. Whether Xerox simply knew it was potentially dangerous, or because they wanted to stay ahead of competitors, Xerox purchased Kurzweil Computer Products in 1980 from Ray Kurzweil. Kurzweil’s company had developed a product that combined omni-font OCR, CCD flatbed scanning, and a text-to-speech synthesizer as the Reading Machine allowing the blind to easily consume textual information. This company became Xerox Imaging Systems, and later Scansoft. By 1985, Xerox had gone from 85% market share in 1974 down to just 40%. The 1970s had been rough. Legal issues, new competitors, and no highly successful new products.

The company had been working to increase reliability and decrease costs, and this kept them alive. But, they finally figured out how to get new technologies to market too. In 1981, Xerox released the Memorywriter electric typewriter. This product was successful, captured more than 20% of the market, and later iterations would add internal memory and other features.

Four years after the Alto was working, one year after the company had a formal method for transferring technology, Xerox began working on a marketable version of the Alto. This work culminated in the Xerox 8010 Star Information System, better known simply as the Xerox Star. This system was built around the AMD Am2900 series with a good amount of microcode implementing an instruction set designed for Mesa, but this could be changed to allow for Interlisp or Smalltalk. The Star shipped with 384K RAM that was expandable to 1.5MB, an HDD of either 10MB, 29MB, or 40MB, an eight inch floppy disk drive, a monochrome seventeen inch display of 1024 by 808 pixels, a two button mouse, ethernet, the Pilot operating system, and the Star desktop software. The system in a base configuration had performance similar to that of the VAX-11/750, and pricing started at $16,000.

The GUI was more advanced than that of the Alto with skeuomorphic desktop icons, scroll bars on windows, and even toolbars with icons indicating functionality. The system was designed around objects where a document for a word-processing program holds page objects, paragraph objects, and so on. These can be selected, and then opened, deleted, copied, or moved in a standardized manner. The settings of any object could also be viewed in a standardized way, and objects could be embedded into any type of document as the objects were independent. Similar technology would show up on the Lisa and on Windows with OLE.

As the microcode could be changed this led to variations as Lisp machines with the 1100, 1132, 1108, 1109, and 1186 Lisp machines. The final variant was the Xerox Daybreak marketed as the 6085 PCS in 1985. By this time, the system shipped with an 80MB HDD, 3.7MB of RAM, a 5.25 inch floppy disk drive, an ethernet controller, and a PC emulator card with an 80186. When this machine was configured with Lisp as opposed to Mesa, it was the 1186. Display options now included either a 15 inch or a 19 inch display.

Another change over time was that the Star desktop had evolved to be ViewPoint which integrated desktop publishing, word processing, spreadsheets, and databases into a rather seamless suite. This would later become GlobalView for IBM-compatible systems running Windows 3.1, Solaris, or OS/2.

The Star series of workstations weren’t successful. The estimates I’ve read state that only around 25000 were sold. Even within Xerox, the Xerox 820 and 820-II systems running CP/M were more common.

If you’re thinking that the Macintosh UI looks similar to the Star, you’re not alone. Xerox thought so too, and thanks to Apple’s suit against Microsoft, they knew that they might have a case. On the 14th of December in 1989, Xerox filed a copyright suit against Apple, though there was the additional claim that Apple’s suit against Microsoft and HP hurt Xerox’s chances of licensing their own technologies. This case was dismissed on the 24th of March in 1990 with Judge Walker having stated that the proper place for this action would by the Copyright Office and not the courtroom, and that in the case of unfair competition, the matter was really a simple copyright infringement case.

In 1982, David T. Kearns took over as CEO. That year, the company launched the 10-series, Marathon, line of copiers which used microprocessors to allow the use of multiple paper types and more complex functions. The Marathon line was the most successful line of Xerox machines, and it allowed the company to start regaining market share. From the company’s product guide:

Available in five configurations so that you can choose exactly the features you need. Common to all are: automatic double-sided copying from single-sided originals or double-sided originals in the case of Systems 1 and 2; reduction for producing standard copies from oversize originals; the copy quality selector and monitoring system which ensures good quality copies, even from difficult originals. And to keep a record of each user’s copying volume, there’s the option of installing a user code system whereby access to the machine is by entering a personal code.

There’s also image shift which ensures that when you’re doing double-sided copying, the image on the second side is correctly positioned, leaving sufficient margin so that no information is lost in the binding. The Xerox 1075 operates at 70 copies a minute and there’s the innovative visual display screen and instruction panel to guide you through all your copying tasks with words and pictures.

System 5 is the basic 1075 model for straightforward, high quality, fast copying. Systems 1-4 incorporate a selection of practical additions to the basic model.

System 1 with fully automatic document handler and finisher is for when you regularly need lengthy reports copied and stapled.

System 2 is ideal for multi-page documents which do not require stapling since, after copying, they are either to be combined with other material or bound.

System 3 has a semi-automatic document handler (SADH), sorter and computer forms feeder. Suitable if your copying does not include a high proportion of multi-page originals.

System 4 gives speed and convenient one-off copying. The SADH has an integral computer forms feeder which positions and copies continuous stationery like computer printout at up to 35 ‘pages’ a minute.

In 1983, Xerox purchased Crum and Forster, a property casualty insurer, and this became Xerox Financial Services in 1984. XFS then purchased some investment firms. By the late 1980s, XFS represented half of all Xerox revenues.

Between 1988 and 1992, Xerox underwent three restructurings. In the first, the company cut around 2000 jobs, dropped its medical systems group, its electronic typewriter group, its workstation group, and created the new Integrated Systems Operations group aiming to shorten time to market. In 1990, Kearns left the company to become the US Deputy Secretary of Education and Paul A. Allaire became CEO. This led to the next two restructuring efforts. The latter of which in February of 1992 led to another reduction in force of 2500. During all of this, the company brought five new printers to market (4135, 4197, 4213, 4235, 4350), the Xerox 5775 Digital Color Copier, and the DocuTech Production Publisher Model 135. The 5775 in particular marked an important change for the industry: a copier was now a microcomputer presented as a scanner and laser printer with a network interface.

Later in 1992, Xerox developed PaperWorks from PARC’s DataGlyphs developed in 1989. DataGlyphs were a type of stealthy 2D bar code that employed steganography. Essentially, DataGlyphs employed an encoding technique where digital data is converted into a pattern of small dots. These dots are integrated into the existing patterns within an image, like the shading or a gradient. The key was that the placement of these dots was controlled, and the data blended into the image such that humans couldn’t perceive it. PaperWorks software utilized this technology to allow fax machines and computers to exchange information, and it was quite clever. The idea was that a computer equipped with a modem could both send and receive faxes. Combining this was DataGlyphs and OCR could then allow a fax to be both input and output for a computer. In the case of the first release of PaperWorks, this meant that a fax could serve as data to be stored or as instructions for specific data to be sent. As cool as this technology was, it was poorly timed with the web revolution just around the corner.

As though giving a rebirth to Xerox’s early copier-as-a-service model, the company did a bit of a rebrand in 1994 as “the Document Company” and they changed their logo to a digital X. The new business model they pursued was to offer Xerox machines, ink, paper, maintenance, configuration, and user support as a subscription service.

The company’s lines of networked digital MFMs (multifunction machine: scanner, copier, printer, fax) became primary revenue drivers while the older xerographic machines declined. By 1997, the MFMs brought in $6.7 billion. The prices of these machines also declined rapidly, and this brought them into homes via stores like CompUSA, Staples, and OfficeMax.

Richard Thoman had joined the company as President and COO in April of 1997 from IBM. His mission had been to once again reinvent Xerox to allow the company to survive; this time not just computers, but the growing World Wide Web. He’d become CEO in April of 1999, and the company’s share price fell by about 50%. Thoman resigned on the 11th of May in 2000. Allaire returned as CEO and Anne Mulcahy took the role of President and COO. The reasons for this rapid fluctuation in the executive ranks was a combination of issues. First, renewed competition from Canon, Ricoh, Ikon, and others ate into market share and therefore revenue. Second, Brazil had a particularly bad economic crisis in earlier in the year, and Brazil was a major market for Xerox. This forced Xerox to reorganize their global sales force where instead of focusing on geography the sales teams would focus on specific market segments or industries. This was needed, but the implementation was costly and poorly executed. Staff were frustrated, customers were angry, and the company’s sales declined. Additionally, the company now had about $16 billion in debt. By the time Allaire returned and Mulcahy was promoted, the company’s share price had fallen another 10% for a total of about 60% from its price a year prior.

In October of 2000, the company posted a quarterly loss of $167 million, and the SEC began an investigation into the company’s accounting practices. This eventually led to a fine of $10 million for booking lease revenue up front. This further led to five company executives being fined individually for a sum totaling $22 million. The shortfall and SEC investigation led to a string of cost cutting measures that would see around 12,400 people laid off by the end of 2001. On the 1st of August in 2001, Mulcahy became CEO. In early 2002, Kearns retired, and Mulcahy additionally became the company’s chairman. Shortly after, PARC became an independent wholly owned subsidiary of Xerox. From January of 2001 to December of 2003, the company’s workforce shrank from 92,500 to 61,100, but under Mulcahy’s leadership, the company launched more than thirty new products, and posted income of $360 million for 2003. By 2004, the company’s debt had been reduced to under $10 billion. Ursula Burns succeeded Mulcahy as CEO on the 1st of July in 2009.

On the 28th of September in 2009, Xerox announced its acquisition of Affiliated Computer Services for $6.9 billion. ACS was an IT services company of around 74,000 people operating in over ninety countries. ACS’s largest business segments were in electronic toll collection, parking systems, photo traffic enforcement, and non-bank-owned ATMs. After becoming part of Xerox, the IT outsourcing arm of ACS failed to grow. That portion of the company was sold to Atos for $1.05 billion. In January of 2016, shareholder Carl Icahn exerted pressure on the company to make the business services unit (mostly ACS) its own publicly traded company. Burns oversaw this separation, and then became the chairman of the new company, Conduent, on the 1st of January in 2017. Jeff Jacobson then became CEO of Xerox.

Jacobson’s tenure was short lived as was chairman Robert Keegan’s. Fujifilm had been looking to acquire 50.1% of Xerox for $6.1 billion and merge all of Xerox into the existing Fuji Xerox joint venture. Jacobson would have become the CEO of the resulting Fuji Xerox, and Xerox’s shareholders would have gained $2.5 billion. This deal was canceled on the 13th of May in 2018, and John Visentin, one of Icahn’s consultants, became CEO. He’d previously held positions at Hewlett-Packard and IBM. The Icahn led board then proceeded to attempt a takeover of HP which failed. That offer was canceled on the 31st of March in 2020.

Steve Bandrowczak became CEO on the 3rd of August in 2022. Xerox PARC was donated to SRI International in April of 2023, but Xerox kept its patent rights. PARC became SRI’s Future Concepts division on the 18th of January in 2024. On the 2nd of July in 2025, Xerox completed the acquisition of Lexmark International for $1.5 billion. For the year of 2024, Xerox’s revenue was $6.22 billion, assets totaled $8.37 billion, and net income was negative $1.3 billion.

Xerox is a company that changed the world more than once. Xerography alone changed the world with easy, low cost, low mess copies. The technology was so successful that Xerox became a verb. PARC’s innovations further transformed the world with EDA, ethernet, and GUIs. While many people say that Xerox was a company that could have been IBM or Microsoft, or some crazy combination of the two, I do not believe this to have been the case. Xerox didn’t have the necessary culture, business practices, and business relationships to achieve the Macintosh or the DeskPro 386. The Alto was too early, too expensive, and lacked a user-friendly design. The Star was too expensive, lacked the user friendliness of its competitors, and it didn’t fit in well with the rest of Xerox’s business. For Xerox to have truly succeeded in the computer business, Xerox would have needed to have transformed itself into a computer company. This was never in the cards. Xerox did well with copiers, and they incubated the future of our world. I’d say that they were plenty successful. Thank you to the fine folks who built, worked at, and supported Xerox over the years. Your contributions to the world are astounding and much appreciated.

My dear readers, many of you worked at, ran, or even founded the companies I cover here on ARF, and some of you were present at those companies for the time periods I cover. A few of you have been mentioned by name. All corrections to the record are sincerely welcome, and I would love any additional insights, corrections, or feedback. Please feel free to leave a comment.