The Altair 8800: The Machine That Launched the PC Revolution

This is part of a series of posts about the circumstances leading up to the launch of the Altair 8800 in the January 1975 issue of Popular Electronics. In my last few posts, I talked about the invention of the concept of the personal computer; the creation of the 4004, the world's first microprocessor; the development of the machines that could claim to be early PCs; and the introduction of the Intel 8080, the processor on which the Altair would be based.

Given the interest in smaller computers, coupled with the introduction of microprocessors, the creation of a successful personal computer was probably inevitable. But it likely would have happened later if it wasn't for a couple of Ziff-Davis editors, who were looking for an attention-grabbing cover story. The January 1975 issue of Popular Electronics, with its cover proclaiming "Project Breakthrough! World's First Minicomputer Kit to Rival Commercial models...Altair 8800" not only grabbed people's attention, but it also proved to be a vital spark in making the personal computer real for a large number of people.

Looking for a Computer Project
The road that led to that point was more than a bit complex. In some ways, it started when Radio-Electronics, a major competitor to Popular Electronics, published a 1973 cover story on Don Lancaster's "TV Typewriter," which allowed readers to buy a kit that would let them display alphanumeric characters, encoded in ASCII, on an ordinary television set.

Arthur Salsberg, editorial director of Popular Electronics, then started looking for a project that would involve the microprocessor and turned to technical editor Leslie "Les" Solomon to create a do-it-yourself computer. "Uncle Sol," as he was called, had created a group of hobbyists and writers who were often creating projects for the magazine to run.

According to Paul Freiberger and Michael Swain's Fire in the Valley, "Solomon and editorial director Arthur Salsberg wanted to publish a piece on building a computer at home. Neither of them knew if such a thing were even possible, but in their bones, they felt that it should be."

By the middle of 1974, as Salsberg described it, "we got in a bunch of competitors," but the search for a computer kit came down to two choices. One was a "computer trainer" designed for hobbyists to teach themselves about computers, based on an Intel 8008 and designed by Jerry Ogdin. The problem with this, Salsberg said, was that the 8008 was "an about-to-be-phased-out chip." The other one, Salsberg said, "amounted to no more than a promise. The promise was, I can get the chips at a lower price and make the whole thing feasible. That was from Ed Roberts."

Henry Edward "Ed" Roberts was the president of a small firm in Albuquerque, New Mexico called Micro Instrumentation and Telemetry Systems (MITS), which originally sold gadgets for radio-controlled airplanes and model rockets. Solomon had been introduced to him by Popular Electronics contributor and MITS co-founder Forest Mims, most likely in the summer of 1971. (Solomon's subsequent tales say 1972, but that's not possible.) In that first meeting, Roberts said he had an idea for a kit for an electronics calculator. In the November 1971 issue of Popular Electronics, a cover story entitled "An Electronic Desk Calculator You Can Build" appeared under Roberts's byline, and the company shifted focus to the calculator market. By the spring of 1974, Roberts was talking about building a microprocessor-based computer.

But that plan was still in the early stages, where Ogdin's was much closer to reality. Solomon and Salsberg's plan was to change once they saw the July 1974 issue of Radio-Electronics with the headline, "Build the Mark-8: Your Personal Minicomputer." The Mark-8 was positioned as a full 8008-based computer, which did more than Ogdin's machine. In most tellings, Salsberg read the story and said "that kills the trainer," leaving the magazine with Roberts's proposal.

There are somewhat different variations on the exact history of the decision to go with Robert's machine.

In Solomon's version, "one of our competitors, Radio-Electronics, was preparing a story on a 'computer' using an Intel 8008 microprocessor. Roberts looked into it, obtained an even newer Intel chip called the 8080, and with a couple of engineering friends set about creating his own computer."

"The MITS computer was ready that summer. Roberts said it could be sold as a kit for about $400, which was fantastic because I knew the Radio-Electronics Mark-8 computer was having difficulties (no peripherals, no language, etc.)." In his recollection, "Art Salsberg, my boss, said he would go along with me on publishing a construction article on a microcomputer ('Heaven only knows who will build one!')."

Salsberg had a somewhat different memory. Responding to Solomon's version, he said "the idea for seeking out a computer for hobbyists started when I read a manuscript submitted by Don Lancaster around January 1974. It described plans to build an ASCII keyboard and encoder for less than $40. I wondered if the main part of a small computer could be made available at a similarly reduced cost. Discussing this with Solomon, I directed him to scout around to see if someone could develop this as a kit, while I would do the same."

In Salsberg's tale, they couldn't find a satisfactory proposal, and instead planned to publish Ogdin's "computer trainer" proposal. "Near the beginning of summer 1974, I think, Radio-Electronics featured a computer project that used an 8008 CPU. It took away our thunder, I felt."

Salsberg says he showed the article to Solomon, "who was not at all aware of its existence." He says he also showed him an article about Intel's newer and more powerful 8080 CPU (which was announced in April 1974) and said they should try to get a computer built around this chip. Salsberg and Solomon then went through a list of possible people to build the computer, and Salsberg asked "isn't there anyone else?"

"Solomon then casually mentioned that Roberts was working on a computer, but that he was not near completing it. I quickly instructed Solomon to telephone Roberts and to dangle a cover story on the computer if he would make our deadline and if the computer was sufficiently powerful. I also told Solomon to tell Roberts that he must include an attractive cabinet with the kit because this would make it more attractive to readers. Solomon returned to my office later and excitedly told me that Roberts said he could make the deadline. He barely did, making history with it."

Building the Altair
In Albuquerque in the spring of 1974, Ed Roberts was running MITS, which by that time had moved most of its focus to calculators. But it was in trouble, because the price of calculators had declined.

Roberts said that by then, MITS was shipping calculators for $35, but it was costing them $33 to build one, leaving no profits. Other similar calculators were retailing for $26 or $28. "They were selling way below what their costs were," Roberts said.

Indeed, MITS was heavily in debt with a bank overdraft of more than $300,000.

"We damn near lost our butt," Roberts said. But it was then that he decided to build a personal computer. "It certainly wasn't done with the idea that it would save MITS," he said. "It was much more a labor of love."

Roberts considered building a machine around the 8008, until a programmer told him it was too slow to be useful. (Some of the folks working on the Micral might have disagreed with this assessment, but no one in the U.S. had seen the machine at this point.)

But when Intel came out with a new chip, the 8080 microprocessor, Roberts called up the company for some horse-trading. Bought in small lots, the chips cost $350 each. But Roberts was not thinking in small lots, so he "beat Intel over the head" to get the chips for $75 apiece by buying the chips in volume.

At this point, he started his discussions with Popular Electronics in earnest. After the Mark-8 story appeared in Radio-Electronics, Solomon flew to Albuquerque to see if Roberts could actually produce a computer for the magazine. He told Roberts that Salsberg wanted the computer to be packaged like a commercial product, not another "rats' nest," and that he wanted it to sell for under $500. Roberts promised to meet the price and to deliver the first machine to Popular Electronics as soon as it was built, and Popular Electronics promised to publish a series of articles on it, including a cover story.

As Freiberger and Swain describe it: "When Salsberg agreed to go with Roberts's machine, he staked the reputation of the magazine on a promise and a hunch. No one at MITS had ever built a computer before. Roberts had only two engineers on the staff, and one of them had his degree in aeronautical engineering. Roberts had no prototype and no detailed proposal. But Uncle Sol kept assuring Salsberg that Roberts could pull it off. Salsberg hoped he was right.

"Roberts was just as edgy about Popular Electronics's promise. However much he liked and respected Les Solomon, he was wary of Solomon's cheerful assurance. The more he realized how important a cover story in popular Electronics was for MITS, the more nervous he became. His company's future was in the hands of a man who levitated tables for kicks."

Roberts wanted to make sure the machine he was building was a full computer. Later, he explained: "The basic ground rules for a personal computer from a technical standpoint is that it had to be a real, fully operational computer that was fully expandable and at least in principal could do anything that a general purpose minicomputer of the time could do. 'Minicomputer' was the term then and referred to any 16-bit or 8-bit machine. And, those were the ground rules. We wanted to make a machine that was, from a user's stand-point, not degenerative at all. The main difference between our machine and where others were is that we used microprocessors, and everything was the latest state of the art. We never used core memory even though we did look at core. At the time we began work on the Altair, core memory was still significantly cheaper than IC based memory."

That summer, the bulk of the machine was designed. MITS only had a small team, with most of the work done by Roberts, senior engineer William "Bill" Yates, and Bybe.

According to MITS co-founder Forest Mims, Roberts designed the interface logic for the 8080, a 256-byte RAM memory, a 2MHz clock, and the front panel logic for the 25 control/input switches and 36 indicator LEDs on the machine; while Yates laid out the foil patterns for the circuit boards.

As Mims describes it, Roberts "also made what was to prove a momentous decision: He included provisions for an open bus, so additional memory and peripheral cards could be added later. The oversize Optima cabinet could accommodate up to 16 additional cards. Therefore, Ed designed a hefty 8-ampere power supply for the machine, having no idea that even this much power would later prove inadequate for the dedicated computer fanatics who stuffed their blue and gray cabinets with peripheral cards."

Yates would actually design a hardware bus that initially used a wire with 100 connections for allowing additional cards to be plugged into the main circuit board. Yates had to work very quickly, so he didn't have time for all the niceties of design, something that would come back to bother board designers later, but it worked.

In many ways, the bus would turn out to be one of the major selling points of the computer. The 100-pin bus would soon become an industry standard, with competitors calling it the S-100 bus (although Roberts would always insist it should be called the "Altair bus").

Design turned out not to be the biggest problem; instead the company was close to being insolvent. Roberts needed a loan of $65,000 to proceed. "I really expected us to get shot down," he said, but somehow he talked the bank into the loan. "I thought maybe we could sell 2,000 in a year."

The Lost Machine
Roberts and Yates wanted the new machine to look like a real computer, so they created a case for their prototype that looked much like the Data General Nova with switches and lights on the front.

Once the prototype of the computer was completed, Roberts shipped the first computer to Solomon via a company called Railway Express. Solomon waited for the machine, but it never arrived. Railway Express had apparently lost the computer and soon afterwards declared bankruptcy.

This left both Popular Electronics and MITS in a precarious position. The magazine had committed to a cover story and now didn't have a machine. Solomon had to look at the schematics and take Roberts's word that the thing worked. And Roberts and MITS didn't have enough time to build a new prototype in time for it to be photographed for the cover.

So faced with no machine, they simply faked it. Yates took a blue about half the size of an air conditioner, added tiny switches and two rows of red LEDs to the front panel, and shipped it off to Solomon. And thus, when the January 1975 issue of Popular Electronics went to press, it featured a flashy cover photo of "an empty metal box masquerading as a computer."

In some accounts, the lost machine gave MITS the opportunity to improve the design.

Roberts had always intended the machine to be expandable. "The goal was to make a machine that in all aspects worked just like a standard minicomputer." By that time, MITS had bought a Data General Nova II, and the Altair would resemble the Data General machine.

According to Paul Ceruzzi, the original prototype had four large circuit boards stacked on top of one another, with a wide ribbon cable carrying 100 lines from one board to another. When working on the new prototype, MITS gave it a deeper cabinet and instead connected the wires to a backplane that carried the signals from one board to another. This allowed other boards beyond the original four. Some say the change was due to the extra time; others say Roberts found a supply of 100-slot connectors at a really good price.

By creating an open bus, the Altair was following the lead of many of minicomputer companies of the time, allowing others to design and market cards for the machine. This bus, which Roberts would always refer to as the Altair Bus but which became standardized as the S-100 bus, would eventually enable later compatible machines. Indeed, companies like IMS (later IMSAI) would do to MITS what the later IBM-compatible machines form companies such as Compaq, HP, and Dell, did to IBM.

By the end of 1974, Solomon had received the working prototype. In his telling, "Ed had sent me another computer by a different route. There I was, in a small office in New York with a metal box marked PE-8 on my desk, and an ASR-33 Teletype as the only way of inputting or displaying instructions and data. Between the front-panel switch start-up routine and the noisy Teletype, I was told to take 'that thing' home, which I suppose made the PE-8 the first workable home computer."

Naming the Altair
One question was what the new machine should be called. As with many things in the story, memories differ.

David Bunnell, a technical writer at MITS who would go on to be the founding editor of PC Magazine and a variety of other magazines, initially suggested that Roberts call the machine "Little Brother." When they sat down to write their story for the magazine, Roberts and Yates would call it the "PE-8," hoping the name would prevent the magazine from scuttling the story. But the editors at Popular Electronics thought it needed something catchier.

Solomon often told the story of how the machine got named, apparently first at a user conference Bunnell organized in Albuquerque, and later repeated in a story by Bunnell and Eddie Curie, another MITS vice president, in an early issue of PC Magazine. This story was repeated in Solomon's own Digital Deli (1984, Workman Publishing Company) and in Fire in the Valley and Steven Levy's Hackers, the two seminal books about the founding of the industry, both published in 1984.

Here's Solomon's version, from his Digital Deli book, which was also excerpted in InfoWorld:

"The next step was finding a catchy name for our 8-bitter. After dinner one night I asked my twelve year-old daughter, who was watching Star Trek, what the computer on the Enterprise was called.
   'Computer,' she answered.
   That's a nice name, I thought, but not sexy. Then she said:
   'Why don't you call it Altair? That's where the Enterprise is going in this episode.'

The next day I called Ed to try out the new name. His answer was curt: 'I don't care what you call it, if we don't sell two hundred we're doomed!' So Altair it became."

It's a fun story and one that has been repeated in many histories of the industry, up through and including Walter Isaacson's The Innovators in 2014.

But Mims tells a different story, saying Solomon discussed the name with Popular Electronics associate editor Alexander Burawa and assistant technical editor John McVeigh. In his account, "Al later remembered saying, 'It's a stellar event, so let's name it after a star.' Within a few minutes, John McVeigh said 'Altair.'"

And Salsberg, replying to the Solomon's InfoWorld account, seems to confirm the later story:

"It is my understanding that it was suggested by John McVeigh, a staff editor, during a meeting with two other staff editors, Solomon and Al Burawa, the latter now managing editor of Modern Electronics, who confirms this. Solomon weaves a nice story about his daughter's naming the machine while watching 'Star Trek,' but it seems that this is just a story."

The Story That Launched the PC Industry
The cover story proclaiming the launch of the Altair 8800 finally appeared in the January 1975 issue of Popular Electronics. Inside, the headline promised, "Exclusive! ALTAIR 8800. The most powerful minicomputer project ever presented—can be built for under $400."

"The era of the computer in every home—a favorite topic among science fiction writers—has arrived!" the story, credited to H. Edward Roberts and William Yates, began. "It's made possible by the Popular Electronics/MITS Altair 8800, a full-blown computer that can hold its own against sophisticated minicomputers now on the market."

The article lists 23 potential applications for the machine, none of them games, although games were to be the first use for most of the buyers. And it promised that readers could order the complete kit—which included the Intel 8080 processor and 256 bytes of memory, for $397, or an assembled version for $498, plus shipping. Given that the list price for an 8080 by itself was $360, this seemed like quite the bargain.

The reaction to the story was immediate. As Roberts recalled, "the first day we got one or two calls. The next day, we may not have gotten a single call, but by the end of the week we could see that we had sold 10 or 15 machines. And we knew at the end of the week that it was big. There was one day late in January or February where we sold 200 machines in one day."

This was followed by another article by Roberts and Yates in the subsequent issue, describing how to program the Altair.

Roberts said that by the middle of January, the company, which a year before had been close to bankruptcy, had cleared its overdraft and had $250,000 in its account.

But the machines still hadn't shipped. According to one story, within a week of the story appearing, MITS had 200 orders, and by the end of February, it had 2,000, "and still all they had was one working prototype." The company managed to ship some board sets by early April; in May, they started shipping complete kits.

MITS had promised 60-day delivery, but orders weren't filled in quantity until the summer, according to Fire in the Valley. And since many of the orders came for the kit, the quality of the final product depended in part on not only the quality of the parts, but also the skill of the hobbyist. By most accounts, it was harder to put together than a typical Heathkit electronics project of the era.

And, of course, the bare machine was pretty limited. With just 256 bytes of memory and no peripherals, the best you could do was make the lights blink in certain patterns.

But it was real and commercially available, and hobbyists were rushing to get the machine.

"One of the things that became evident right away was that people lusted after those machines. I don't know any better way to describe it: They lusted after it," Ed Roberts told W. David Gardner in Reflections: An Oral History of the Computer Industry, a supplement to Computer Systems News.

On a chilly December day in Harvard Square, Paul Allen saw a copy of the January issue of Popular Electronics, which "stopped me in my tracks," he wrote in Idea Man. He bought the issue and showed it to his friend Bill Gates, and together they made the decision to write a version of BASIC for the machine. (This was actually written on a PDP-10 minicomputer at Harvard, using an 8080 simulator, before either man had actually seen the Altair itself.) Allen brought the program to Roberts, and "director of software development" at MITS, where he and Gates would work on software. This eventually led to a lawsuit between the pair and MITS over the ownership of the BASIC compiler. Gates and Allen won, and of course the two went on to start Microsoft.

Meanwhile, MITS started work on its own peripheral cards, including interfaces to peripherals and more memory, which the machine desperately needed. "We made a commitment to do a system design," Roberts said. "Before we even talked to Popular Electronics, we had cursory interface for disks, tape drives, and a couple of different types of printers." (Paul Freiberger, Ed Roberts: The Father of the Personal Computer, Popular Computing January 1985, pg 74-79)

Other companies, starting with Process Technology, soon began creating plug-in boards, such as more memory, a way to connect it to a teletype, and later ways of hooking it to a television set and a keyboard, Paul E. Ceruzzi wrote in A History of Modern Computing.

The machine continued to attract a lot of attention in the hobbyist community. On April 16, Steve Dompier reported on MITS at the Homebrew Computing Club in Menlo Park, CA, saying MITS had 4,000 orders at that point, according to Fire in the Valley.

A bit later, the People's Computer Company devoted a page to the new machine, urging readers to get a hold of the Popular Electronics article. Lee Felsenstein (who would later go on to design the Osborne 1) and Bob March read the story and started to build boards for it. This would turn into Processor Technology Company in Berkeley, CA.

According to Steven Levy, Felsenstein "knew that the significance of the Altair was not as a technological advance, or even as a useful product. The value would be in the price and the promise—both of which would entice people to order kits and build their own computers."

Homebrew Computer Club members would start to create machines of their own, including Felsenstein and of course Homebrew member Steve Wozniak, who would soon build his Apple I.

By the end of 1975, you could build a CPU with add-in boards for a little over $1,000, and you could attach a terminal and printer, run Basic, Assembler, and a Debugger. And also by that time, there were direct competitors, including Imsai Manufacturing Corp., which had its own machine based on the 8080 and a compatible bus, which it would call "S-100."

As for MITS, it did $1 million in sales in 1975 and tripled that in 1976. "Our market was bigger than our capacity to expand" Roberts says. When he sold the company in May 1977 to Pertec, Roberts says it was doing $20 million in business annually. Pertec proved unable to market the Altair line in an increasingly competitive industry and discontinued the line a couple of years later. Roberts moved to Georgia to become a doctor, passing away in 2010.

But the impact of the Altair lives on—it was the machine that convinced a large number of hobbyists that the time was right for a "personal computer" and that here was a machine they could actually own. Over the life of the Altair, it sold tens of thousands of units, making it the first volume commercial microcomputers. The computer revolution was on.

For more information, see "Altair Minicomputer, Part I" by H. Edward Roberts and William Yates (Popular Electronics, January 1975), "The Altair story: early days at MITS" by Forrest M. Mims III (Creative Computing, November 1984), "Build the Altair 8800 Minicomputer, Part Two" by H. Edward Roberts and William Yates (Popular Electronics, February 1975), "Build the Mark-8: Your Personal Minicomputer" (Radio-Electronics, July 1974), Digital Deli by Lunch Group (1984, Workman Publishing), "Ed Roberts: The Father of the Personal Computer" by Paul Freiberger (Popular Computing, January 1985), "An Electronic Desk Calculator You Can Build" (Popular Electronics, November 1971), Fire in the Valley by Paul Freiberger and Michael Swain (1984 and 2000, McGraw-Hill), "The Great and Glorious Crusade" by David Bunnell and Eddie Curie (PC Magazine, August 1982), Hackers  by Paul E. Ceruzzi (2003, The MIT Press), The Innovators by Walter Isaacson (2014, Simon & Schuster), "Interview with Ed Roberts" (Historically Brewed Magazine), "Jaded Memory" by Arthur Salsberg (InfoWorld, November 12, 1984), "Les Solomon: favorite uncle & mentor of microcomputing" by Paul Freiberger and Michael Swain (Infoworld, November 29, 1982), "Reflections: An Oral History of the Computer Industry" (supplement to Computer Systems News, 10th anniversary), "Solomon's Memory," (InfoWorld, Oct. 15, 1984).