BOULDER CREEK, Calif.--I have seen the world's first microcomputer, and it is not the Altair.
For years, any serious discussions about the earliest microcomputers had to include the Altair 8800, the creation of Albuquerque, N.M.'s Micro Instrumentation and Telemetry Systems (MITS). That computer, as has been well chronicled, inspired legions of hobbyists, including Bill Gates and Paul Allen, who, upon seeing the Altair on the cover of the January 1975 issue of Popular Electronics, began a mad rush to create Microsoft BASIC, their first smash hit and the beginning of their empire.
But it turns out the Altair was years behind another microcomputer, one that might have turned the computing world on its ear if a few things had gone differently. And while the Altair has been in the public's eye for all these years, its rival for the title of "earliest microcomputer" had been all but lost to history.
It still exists, however. It's in a plastic box at the DigiBarn Computer Museum here, and the founder of that wonderful collection, Bruce Damer, took the time earlier this week to pull it out and show it to me.
The first fully built-out microcomputer
Damer said that about a year-and-a-half ago, he was contacted by a former Sacramento State researcher named Bill Pentz who told him something along the lines of, "I think I led a team at Sacramento State which built the first fully-built-out microcomputer."
Until more recently, that's as far as the story went. But then, a former university worker found the computer's parts in his basement, and Pentz subsequently donated them to the DigiBarn. It's no longer in any kind of assembled shape, but thankfully, the parts are still intact, and could, in theory, be put together again.
The project began in 1972 at Sacramento State University where Pentz's team set out to build a computer capable of handling thousands of patients' medical records.
To do so, explained Damer, who learned the history directly from Pentz, the team began with an Intel 8008, the world's first 8-bit microprocessor, which was built onto a motherboard put together and donated by Oregon's Tektronix, which "took pity on the Sacramento State team," Damer said.
What makes the machine so noteworthy is how complete it was back in 1972 when it was assembled. All together, it had a full set of hardware and software components: a disk operating system included in a series of programmable read-only memory chips (PROMs); 8 Kilobytes of RAM; an assembly language; a hard drive; a color display; a printer output; a 150bps serial interface for connecting to a mainframe; and even the world's first microcomputer front panel.
According to Damer, what made this computer, which he has nicknamed "the Sac State 8008," special was that it was far more sophisticated, even in 1972, than was the Altair 8800 when it was released in 1975. In particular, he noted, the Sacramento State computer had the operating system and the assembly language--in this case, IBM's Basic Assembly Language (BAL).
To be sure, the external components were not built for the computer. Rather, Damer explained, the Sacramento State team pulled them off of a mainframe. Still, the peripherals, such as the hard drive--a Memorex with 3 Megabyte platters--worked with the machine.
"There wasn't a microcomputer equivalent (to this)," Damer said, "until the late 1970s or early 1980s."
Didn't change the world
One might reasonably ask why, if the Sac State 8008 was such an advanced machine for its time, no one knows about it?
The reason, Damer concluded, is that it would have been extremely expensive to replicate, largely because of the PROMs. That, plus the fact that the machine was built in a public university lab for a specific purpose, and not by private hobbyists looking to make a splash, likely led to its anonymity and eventual disappearance into someone's dusty basement.
But there's little doubting its significance in the annals of vintage computer history.
To Lee Felsenstein, an original member of the famed Homebrew Computer Club and the designer of the famous Osborne 1 computer (see video below of Felsenstein and VisiCalc co-creator Bob Frankston discussing the Sac State 8008), there's little question that Pentz's machine was, if not the world's first "personal computer," then at least the world's inaugural complete microcomputer.
Damer suggested that had it not been for the high cost of the Sac State 8008 machine and the public nature of its assembly, it might have gone on to have a much more significant impact: he said that it's possible to imagine that, because of the complete nature of the machine, including its built-in Basic Assembly Language, Gates and Allen might not have had such an urgent need to create Microsoft BASIC. Similarly, because the machine had a disk operating system (DOS), which was compatible with IBM computers, Big Blue might not have gone in search of a DOS for its IBM PC.
Without those two events, there might be no Microsoft today, Damer argued.
That, of course, is a speculative argument, and one that Felsenstein doesn't agree with. But it's certainly interesting to ponder the question of what might have happened had Pentz's computer been cheaper to build or been put together by private hobbyists looking to make a buck.
Either way, however, there's no way to deny that what Pentz's computer did was demonstrate what was possible in microcomputers, several years before the Altair came along.
Still, while acknowledging the true innovation that the Sacramento State computer represented, Felsenstein said that even if it had gotten out into the public eye, it probably wouldn't have altered later events much. In part, he said, that's because hobbyists were looking for a computer they could mess around with and modify to their hearts' content. And that's why the Altair, which came either as a kit or as an assembled machine with nothing but memory, became such a hit.
In other words, Felsenstein said, "the bomb (Microsoft and the explosion of microcomputers after the Altair) was still going to go off."
The earliest mint condition Altair
While I was at the DigiBarn, Damer also showed me another piece of vintage computing treasure.
In this case, it was what is likely the oldest mint condition Altair anywhere on Earth (see video below).
Not too long ago, he said, the computer's owner donated it, along with several other machines, to the DigiBarn. And it wasn't until Damer opened the package that he realized what he had on his hands: Altair 8800 serial number 220047A.
Since all Altair serial numbers began with "22," Damer concluded that this machine was the 47th assembled computer off the production line. The computers were sold assembled, in a box, and those machines' serial numbers ended in "A," or as a kit, with serial numbers ending in "K."
"Unless somebody comes out and says, 'Oh, I've got model 5 with not a scratch on it,'" Damer said, this machine could well be the earliest one that had never before been taken out of its box.
Indeed, the machine is in stellar shape: it looks brand new, and had never once had its case opened since its manufacture in 1975. Until, that is, Damer pulled out a screwdriver and popped the case off in front of me.
There, before us, was a virgin Altair, with nothing but its memory board and several empty slots. It looked new. This could well have been 1975.
Why someone had spent $1,000 back when Gerald Ford was president to buy a computer, only to leave it untouched in its box is a question we may never know the answer to. But for those who appreciate things like vintage computers, this was a truly a special moment.
Correction: This story mistakenly referred to the Sacramento State 8008 system as having a 9,600-baud modem. In fact, it had a serial interface with settings going up to 9,600bps but the mainframe it was attached to could only connect to the microcomputer at 150bps.