The Microsoft Softcard, released April 2, 1980, was Microsoft’s first hardware product. Not only that, it was a hardware product that made two products from two of its historical rivals work together. For those reasons and more, it is a historically very important product. This hardware product from a software company changed Microsoft’s trajectory dramatically.
The Microsoft Softcard was a plug-in circuit board for the Apple II series of computers to add a Z80 processor to make it possible to run CP/M software.
From today’s perspective, it’s understandable to question why Microsoft wanted consumers to be able to run an operating system made by Digital Research on computers made by Apple. It was definitely an example of Paul Allen’s out-of-the-box thinking.
Microsoft and the 6502 processor
Microsoft had a large library of software at the time, mostly programming languages. But Microsoft was, to say the least, not a huge fan of the 6502 processor at the heart of the Apple II.
That’s not to say they didn’t write any native software for the Apple II in 6502 assembly language. They did write an implementation of Microsoft Basic for the 6502, including a specialized version for the Apple II, and they did produce a few software titles for the Apple II, mainly consumer oriented titles rather than programming languages.
But there are other examples of titles that they intentionally stayed away from on the Apple II and other 6502 based machines. Microsoft Multiplan, their spreadsheet before Excel, was available on 6502 machines but only because Microsoft licensed the technology to somebody else. They were fine with someone else doing the port, but they didn’t want to do it themselves. And when they approached a firm called subLogic about licensing the technology that became Microsoft Flight Simulator, they only wanted it on the IBM PC. They didn’t get involved in subLogic’s flight simulators for 6502-based platforms.
So when it came to their large collection of software like programming languages and the Multiplan spreadsheet, Paul Allen’s idea was to create a hardware product to bring the Apple II to them, rather than the other way around.
The hardware behind the Microsoft Softcard
The Microsoft Softcard was a straightforward design, consisting of a Z80 CPU running at 2 MHz and 74-series TTL logic chips, all off-the-shelf stuff. It sold for $349.
A revised version from 1983 added 80 column capability, 16K of RAM, and mixed case characters, so it could replace three other add-in cards and provided some benefit when you were running Apple II software as well. But the main selling point behind both versions was that you could insert a floppy disk, reboot, and have the Z80 take over and turn the Apple into a CP/M machine that could run hundreds of software titles from Microsoft and other publishers.
At that point in time, the Atari 800 had the largest software library of any microcomputer available. But an Apple II with the Softcard could run Apple II and CP/M software, letting it leapfrog Atari.
The resulting machine didn’t have anywhere near the graphics and sound capability of an Atari 800, but it had a larger software library, both in terms of number of titles and variety of titles.
Of course Microsoft hoped the Softcard would be a success. It ended up far exceeding their sales expectations, selling 5,000 units in its first three months and becoming Microsoft’s leading revenue source–for a time.
The unexpected historical significance of the Microsoft Softcard
The Softcard was more successful than Microsoft expected. But it also accidentally changed the course of history. Ironically, it was designed by Tim Paterson. If his name sounds familiar, it’s going to come up again. Early Microsoft history is a tangled web.
When IBM decided to start designing its personal computer, they knew that they wanted Microsoft Basic and the CP/M operating system. Crucially, they talked to Microsoft first, possibly because IBM’s CEO knew Bill Gates’ mother. IBM first approached Microsoft in July 1980. Because they knew Microsoft was supplying CP/M for Apple machines, they hoped they could buy both from Microsoft rather than dealing with multiple vendors. At least that’s how the common version of the story goes.
Microsoft was happy to provide Basic for the IBM PC and even to collaborate with IBM on the overall design of the machine. But Microsoft couldn’t sell them CP/M. Microsoft had an operating system to sell them, a Unix variant called Xenix. But Xenix wouldn’t run on the Intel 8088 CPU IBM wanted to use. If IBM would consider the Motorola 68000, Microsoft could supply them with an operating system and Microsoft Basic.
You read that right. Bill Gates wanted the IBM PC to be a Unix workstation that ran on the same CPU that ended up in the Apple Macintosh.
But IBM had already considered the 68000 and determined Motorola couldn’t meet their aggressive timeline. Besides, they preferred the 8088 and CP/M, a combination they thought would sell well while presenting no threat to sales of their more powerful and expensive minicomputers and mainframes. This mindset was a recurring theme for IBM in the 1980s. But that’s another story. Make that two stories.
Since IBM had its mind made up on CP/M, Gates referred them to Digital Research.
IBM returns to Microsoft and Microsoft calls up an old friend
IBM’s talks with Digital Research went badly. IBM ended up coming back to Microsoft, who ended up supplying IBM with not only Basic but also an operating system. Remember Tim Paterson, designer of the Microsoft Softcard? Paterson wasn’t just a hardware designer. He could program too, and had developed a CP/M clone that happened to run on the 8088 CPU that IBM wanted to use. Microsoft licensed Paterson’s operating system, then turned around and licensed it to IBM. You now know that operating system as PC DOS, or MS-DOS. Gary Kildall, the founder of Digital Research and original author of CP/M, was bitter for the rest of his life.
That whole story, which has attained legendary status, could have gone very differently if Microsoft hadn’t been selling the Softcard. IBM would have had no reason to ask Microsoft about CP/M in that case. IBM might very well have approached Digital Research first, and the timing certainly would have been different. And a lot of the problem with the misunderstanding between IBM and Digital Research was simply bad timing.
So there is more than a little bit of irony that a hardware product that Microsoft made for Apple computers led directly to Microsoft and Apple becoming rivals.