What ever happened to…
MSX computers
by John C. Dvorak
Not too many people will recall the short-lived era of the “MSX” initiative which was slated to pretty much take over the non-existent middle world where consumer electronics met personal computers. It was always believed, back then, that this is where the sweet spot of profits would emerge. What emerged was instead laughable MSX. It was one of Microsoft’s greatest flops.
In the second half of 1982 Microsoft, working with ASCII Corp. of Japan. developed a new hardware and software standard, MSX or Microsoft Extended, that would use “proven technology” to produce a cheap 8-bit home computer. Although Commodore, Atari, Mattel, Coleco and many others were already in the U.S. market selling cheap home computers, Microsoft, taking a cue from IBM and it’s “standardized PC,” saw a need for standards as an opportunity to wipe out the competition.
With its Japanese partner and distributor, ASCII Corp., Microsoft convinced some of Japan’s largest consumer electronics and computer manufacturers to license the technology to build MSX machines. The names were big and were also the “usual suspects”: Sony, Toshiba, Pioneer, Hitachi, and Yamaha, to name a few.
It was ASCII’s cofounder, Kazuhiko “Kaye” Nishi, who was the driving force behind the acceptance of the MSX standard in Japan. Nishi and he had been a strong supporter of Microsoft’s since the late 70’s. It was ASCII who eventually sold Yamaha MSX machines to the Soviet Union and even developed its own 16-bit 28 Mhz version of the Z-80 chip, dubbed the R800, to be used in the 16-bit version of the MSX system. The design for the chip was eventually licensed to Zilog, originators of the Z-80.
Nishi was also credited with perhaps inventing the laptop computer.
It was back in 1983, though, when things looked rosy for MSX. Following Microsoft’s official Japanese announcement in June, 1983, a half dozen MSX computers were introduced at the Japan Data Show in October. Retailing in the $200- $400 price range, not including a monitor, more than 265,000 units from about a dozen manufacturers had been sold by midsummer of 1984. The MSX standard also showed signs of success in Europe, where electronics giant Philips was its main proponent. There was an MSX-specific magazine introduced in England to take advantage of the predicted boom. Still, the reception in the US was negligible.
The MSX standard was built around the 4Mhz, 8-bit Zilog Z80A CPU, and included other popular, and cheap, components such as the Texas Instruments TMS9918A video chip and the General Instruments AY-3-8910 programmable sound chip. Standard configurations included 16K, 32K or 64K of RAM, BASIC and BIOS in 32K of ROM, and 16K of video RAM. And its version of BASIC, MSX BASIC, was excellent; providing easy programming access to the computer’s sound and video capabilities.
For a low-end computer, the MSX standard was amazingly efficient. Its performance was on a par with the Apple IIc, and the IBM PCjr. In fact IBM was reportedly taking a long look at the MSX standard around the time it was being introduced to American manufacturers.
Unfortunately, it was not a disk-based system. Until MSX-DOS was introduced in 1984, MSX computers used only cassette and cartridge interfaces to load programs. The US market had long since showed a distinct inclination towards disk-based machines. This MSX idea was a throwback to 1977.
That stumble combined with a shake-out in the American home computer market, made the Japanese manufacturers take a wait-and-see attitude for much of 1984. Although Microsoft hosted an exclusive preview for third-party developers at the June, 1984 Consumer Electronics Show, the main Japanese manufacturers didn’t introduce their products in the U.S. until the January, 1985 show. By then it was too late.
At the 1984 CES Atari introduced their new ST line and Commodore introduced their 128 model. The ST and the 128 were priced at the high-end of the MSX market, which meant that in order to compete in the U.S. market the price of MSX computers would have to be slashed. And the call for 16 and 32-bit computing made the consumer aware of the fact that 8-bit was dead. Wisely, no American manufacturers could be enticed to produce an MSX computer. And the fear of fighting the Japanese in a commodity market with rapidly declining profit margins was also a major deterrent. With no commitment from major American manufacturers, none of the software houses were willing to make a commitment either.
In a move to go more upscale and to 16-bit the MSX II (1985) and MSX II-Turbo (1988) versions of the standard were eventually developed. In 1985 Panasonic said it would develop an upscale business machine using the MSX II standard. Toshiba said it was planning a single chip z-80/MSX machine. These announcements met with a lukewarm reception. By the time the “turbo” machine standard rolled out virtually all makers of MSX equipment had stopped development.
Shortly after the MSX fiasco Microsoft and ASCII Corp. parted company in a disagreement over the direction for future developments in Japan. Microsoft ended the exclusive distributor agreement they had with ASCII and opened a Japanese subsidiary.
In February of 1990 one of the last hold outs for MSX, Sony Corporation, gave up on supporting MSX and moved into production of AX computers (another dead-end) and mini-notebook machines. Only Panasonic continued to make an MSX-II Turbo machine. Sanyo also stayed in the consortium to no avail
The story of MSX – if it is ever remembered — might eventually be a Harvard Business School case study. The entire episode proves that popularizing a new standard is not a trivial pursuit.