PC MEMORIES, HOW I CREATED THE PC
A Conversation with David Bunnell [3.13.00]
A long, long time ago, way back in 1973, the world was slow and dreary and mostly analog. There was no Federal Express, no e-mail and no venture-capitalist empire on Sand Hill Road. Where there are now chip-fabrication plants and pure business-to-business Internet plays, peach and cherry orchards abounded. Steve Jobs was barefoot in India, and Marc Andreessen was just learning to crawl. Some businesses had fax machines, but the faxes were printed out on expensive, smelly paper that smeared the ink. No one knew what a spreadsheet was. There were no pagers, no cell phones, no laptop computers. People actually wrote letters and even books on IBM Selectric typewriters.
Gordon Moore and Robert Noyce had founded Intel, but no one had heard or cared about it: Chips were just not yet all that glamorous. William Shockley had quit making microprocessors and was preaching genetics at Stanford University. Electronic calculators were the newest thing, if you could afford one, but most engineering students still had slide rules hanging from their belts. Down in Atlanta, Ted Turner was hawking billboard advertising. Most people still had black-and-white TV sets that got four channels: ABC, NBC, CBS and PBS. And everyone, it seemed, was watching the Watergate hearings.
Me? I was an ambitious 24-year-old looking for a way to get out of Albuquerque. And although I didn't know it yet, I had just landed the most important job of my life as a technical writer at a "micro-electronics" company on the East Mesa called MITS. I didn't know much about electronics, but the very word sounded like something that could be really big. So I was very excited about this job because for the first time I could make a living by doing what I liked best: writing.
Welcome to MITS
I got the job at MITS not because I was qualified (I wasn't), but because the founder of the company, who seemed to have an endless capacity for hobbies and special interests, happened to share my passion for reading about World War II.
I found the job listed in the Albuquerque Tribune and somehow bullshitted my way past two rounds of interviews—one with the "head" technical writer, a nice but much-too-serious woman named Belinda Wilson; the other a brilliant but painfully shy vice president of engineering named Bill Yates (no connection to you know who). It was obvious that Yates would rather have been poring over a schematic diagram, doing the real work of this world, but he seemed somewhat resigned to the bureaucratic functions of his job.
Note: This story is from a fictional autobiography that David Bunnell is writing. Bunnell was actually in Albuquerque working for MITS at the time the Altair was introduced, and he did work with Paul Allen and Bill Gates. Much of this autobiography is true, and much of it is serious embellishment.
Having nothing much to lose, I decided to be pretty straightforward about what I wanted. "Look, Mr. Yates," I said. "I may not know much about electronics, but I have a good background in math and science. In my freshman year at college I was the only student to get an A+ in calculus. I can write, I can think and I like to work hard, so please give me a chance. You won't be sorry. I'm the best person for this job."
Yates wasn't sure I was truly qualified, but I think he liked my chutzpah. Besides, there were no other applicants for the position. Also, there was a fairly good chance that if he had turned me down, I would simply have punched his lights out. So he said, "OK, if you can convince the big guy, Ed, that you can do this job, then I guess you can have it. But it's only on a trial basis." And thus Yates led me down MITS' grungy linoleum-tile hallway to Ed Roberts' "suite," which included an alcove for his assistant—a pleasant woman named Barbara who turned out to be the only person at the company who wasn't stressed out 95-plus percent of the time—as well as Ed's rather imposing office and a private bathroom.
I waited in the alcove for a few minutes, chatting with Barbara about mundane things like which restaurant in town had the best blue-corn enchilada and why so many people in Albuquerque wear so much turquoise jewelry. I learned that she had recently graduated from New Mexico State University and was married to someone named Bobby who sold electronic parts for a company called Hamilton, which turned out to be a major supplier to MITS. It all sounded nice and tidy.
Finally, a gigantic man with a crewcut appeared at the doorway and asked, "Are you David Bunnell, the guy who thinks he can be a technical writer?"
"Yup, that's me," I replied. "And you must be Ed Roberts. It's a pleasure to meet you."
I followed the CEO into his office, which smelled of stale cigarette smoke, and sat in front of his gigantic desk, the most prominent accessory of which was a huge circular ashtray filled with about a thousand cigarette butts. Ed sat back in his chair, coughed a couple times and lit one of the bigger butts from the ashtray. "Do you smoke?" he asked. "No, but go ahead," I answered.
Ed began the conversation with a brief history of MITS. The acronym stood for Micro Instrumentation and Telemetry Systems, which reflected the fact that its first products were backyard model rockets. These rockets would blast off into the atmosphere to a height of several thousand feet and return to Earth with a parachute that sometimes opened and sometimes didn't. Ed actually had started the company while he was stationed at Kirkland Air Force Base in Albuquerque. Later I learned from his best friend, Eddie Currie (who also worked at MITS), that even though Ed claimed that MITS was a hobby, it was really the other way around: The Air Force was Ed's hobby, while MITS was his main focus.
The company was currently making build-it-yourself electronic calculators. For a couple hundred bucks you got a kit that included a bunch of transistors and capacitors, a circuit board and other electronic stuff that you soldered together yourself and fit into a plastic case. With any luck, you ended up with your very own fully programmable calculator, and it was a real bargain at the time because HP was selling models with equivalent functions for $900.
Following the history lesson, I told Ed that I had majored in history in college. When he asked me what I specialized in, I replied, "World War II," and we then spent the next six hours talking about that subject. It was after 9 p.m. when I finally stumbled out of his office, my head spinning madly from all the smoke. Yates was waiting for me in the alcove—Barbara had gone home.
"Well," Yates said, "did you get the job?"
"Yes, thanks," I answered.
"And what did you guys talk about for all that time?"
"Well, mostly we discussed the last days in Hitler's bunker. It was great fun, and I think your man likes me."
Seeking My Riches
Technical writing was a dumb job, it turned out, but it quickly became the first step up the MITS corporate ladder. My boss, Belinda, left the company to move to Kansas with her husband, and I was promoted to head technical writer. Soon I found myself producing all of MITS' marketing and advertising materials. Before the first year was out, I became VP of marketing. With each promotion, Ed gave me what I then thought were unbelievably big raises—$10,000, $25,000, $50,000—and before long I was making far more money than Yates, who was jealous of my rise. Ed Roberts, at 300-plus pounds, and I, at 125 pounds, were the Mutt and Jeff of MITS.
Ed was constantly creating new products for his company's hobbyist customers. He had a very active imagination and seemed to know what they wanted. From my point of view, though, these products—such as the MITS 57B Wave-Form Generator—were aimed at very narrow markets of superhobbyists, and no one would get rich by selling them. And it wasn't very likely that the "CBS Evening News" was going to announce that MITS had just come out with a cheaper programmable calculator.
The more I knew Ed and realized how talented but limited he was in his vision of the universe, the more I wanted him to do something big—really big. I, of course, had no real way of judging Ed's engineering skills, although he talked about electronics with the same passion he discussed Hermann Goering. So I had complete trust that he would rise to the challenge. If only I could come up with the vision and then talk him into believing that it was his idea, Ed Roberts could probably figure out how to implement it.
I became obsessed. I wanted nothing less than for Ed to do something that could best be described as "watershed invention," a development that would change the course of mankind. Something bigger than electricity itself.
At the time, MITS employed about 50 people and was located in a tiny shopping center that took up half a block on Linn Avenue just off Route 66 (also called Central Avenue) in a seedy section of Albuquerque. The offices weren't much to look at. Its glory days long past, the shopping center had become home to second-tier businesses like the massage parlor next door and the wholesale office-supply company on the corner. The other half of the block housed a branch office of the Bank of Albuquerque, which was robbed about once every six weeks, a discount paint store, and a Dairy Queen, the lunch spot of choice for MITS employees.
Just across the street was the rundown Sundowner Motel, future home to Microsoft cofounders Bill Gates and Paul Allen. During the day, the whole neighborhood was drab and polluted from all the traffic on Central Avenue, but at night it livened up as a major hangout for drug dealers, prostitutes and their clients.
My previous experiences—which included being fired from several teaching jobs—had given me the burning ambition make myself a rich and powerful guy. However, this was never going to happen as long as I was stuck selling weird electronic toys to part-time rocket scientists. I was ambitious to get out of New Mexico, to strike it rich, to move to California and live on the beach (in grand style, of course).
The Eureka Moment
I was spending several nights a month at the local public library reading through technical trade journals and textbooks, where I learned about basic circuit design, the latest semiconductor research and many other things. My intention was to continue fooling Bill Yates, Ed Roberts and Eddie Currie into believing that I actually did know a great deal about electronics.
I was pretty sure that chips were going to change the world. Everyone was saying so, but everyone also seemed a little vague on the specifics of how this would actually happen. I very much wanted to know how chips would change the world--I felt very strongly that someone would figure it out, and that it might as well be me.
My epiphany came while I was looking at microfiched back issues of Scientific American. I came across an article penned by a nerdy Xerox scientist named Alan Kay. The article discussed some experiments for which Kay had built a prototype "personal computer" called the Alto that used a mouselike pointing device and a keyboard to communicate through a connected video screen. The great, unbelievable thing about this was that no one at that point had commercialized the idea because each Alto machine cost a few hundred thousand dollars to build. And Xerox was a bit lame in any case.
My vision started to take shape: As chips got cheaper and faster and could hold more memory, the day would come when we could build a true personal computer—one that was affordable to most people.
Very shortly after having these revolutionary thoughts, I overheard Ed talking to Currie and Yates about a new 8-bit microprocessor from Intel. This was it. I stormed into Ed's office and shouted, "Damn it, Ed! Why don't we do something that's really big so we can have lots of customers, make lots of money and help people at the same time? Why don't we use this new Intel chip to make a computer kit that hobbyists can afford to buy and build? Just think of it: We'll be putting the power of a mainframe into the hands of the average nerdy guy, who will in turn invent ways for ordinary people to use computers. And then we'll all be rich."
Currie and Yates looked at me with their mouths hanging open while Ed lit up a cigarette and absentmindedly flipped through an issue of Ham Radio Electronics. Ed had already chewed my head off on more than one occasion, so I recognized the risk I was taking—particularly because Ed Roberts viewed himself more as the commander of MITS than as its CEO. But my idea had me so excited that I didn't care. It was, after all, such a compelling concept.
So I just rattled on. "If you build the computer around the 8080 chip, it could be programmed to mimic the functions of a Data General minicomputer, which means it could run Basic. Hobbyists are bound to love Basic. There could be computer clubs, computer stores, computer magazines, computer conventions, computer TV shows—and all of it would spring up if only someone like you could be convinced to build this thing. And the best thing of all is that you could sell it cheap."
I needed Ed Roberts because, when it came to actually building something, I knew nothing about electronics. I had long, hallucinatory dreams about what the world could be like thanks to affordable electronic products aimed at mass markets. But I had no idea about how to build the machines.
Luckily for me, Ed didn't kick the shit out of me just then. Perhaps I'd caught him at a weak moment, because he simply looked up from his magazine and waited patiently with a curious grin as I talked nonstop for about 20 minutes extolling the great history-changing virtues of the personal computer I wanted him to build. Ed coughed a few times—in fact, he nearly choked once or twice—but I talked on.
I told him how we could cut a deal with Intel because, if our invention worked, the chip maker would become the main supplier of microprocessors to the 500-million-person market for personal computers, and this would make Intel the biggest and most profitable chip company on the planet. So why wouldn't Intel give MITS a discount that would give us a chance to prove the concept?
Just as I was finishing my spiel, Ed place his index finger on his lip to tell me to stop talking, and he stared blankly at me for a few moments that seemed like days. I stood in a cold sweat wondering how I would break the news that I'd been fired to my wife, who was already annoyed at me because I seldom came home before midnight. I also wondered where else in Albuquerque I could find a job other than selling used cars.
Finally, Ed gathered his thoughts and, clearing his throat one more time, looked steely eyed right into my eyes. "Damn, Bunnell," he said. "That's one kick-ass idea. We'll get started right away. But you've got to agree never to tell anyone that this was your idea. I want people to think that it was mine."
I had to laugh. Ed wanted to be Father of the Personal Computer. Did I care? Not really. I just wanted to change the world and get rich in the process.
"OK, Ed," I replied. "It's a deal."
Currie and Yates chimed in simultaneously, "Yeah, Ed, sure, you can be the Father of the Personal Computer."
The PC Takes Shape
Ed and I had a wonderful time over the next few months as we spent night after night at MITS, mostly in his office, designing circuit boards and looking for the perfect memory-chip solution. I would usually go home around 4 a.m., fall into bed and sleep until 7 or when the kids woke up, whichever came first.
Taking care of the kids in the morning before they went off to school was my only family time. Ed and I were so involved with our project that we even worked through the weekends, and sometimes when he'd had a big brainstorm, he would call me at 5 a.m. to discuss it.
What I realized only too late was that although Ed Roberts wasn't as intelligent as I was, he was still an awfully bright and deep-thinking man who, unfortunately, had his own ideas. Before long Ed actually started to believe he had invented the personal computer, and instead of merely following my instructions on building the personal computer, which he christened the Altair, he went off on his own tangents.
We argued about everything, and I worried that he would blow the whole deal. His attachment to dynamic memory chips, which were obviously too new and not yet reliable, was one of my biggest concerns. "Damn it, Ed, static memory is more dependable than dynamic memory!" I would scream at him. "Why don't we give the Altair a bigger power supply, just in case we have to switch to static memory chips? I don't know what you have against big power supplies."
Ed refused to give the Altair a big power supply because he was so sure that dynamic memory would be reliable, which turned out not to be the case. The failure rate of dynamic chips was almost 60 percent, so the first Altair had problems right away—and that opened up the market to Altair clones like the dreaded Imsai and Processor Technology computers. Funny how a technical mistake opened us up to competition and set a market pattern that would be repeated several times by other companies as would-be leaders opened up opportunities for someone else to enter their markets. But I warned Ed!
To kick-start the launch of the Altair, I had arranged with Les Solomon, the editor of Popular Electronics, to put the machine on the cover with the headline "World's Cheapest Minicomputer" because I felt that every one of the magazine's readers knew what this meant and that hardly anyone had yet heard of a personal computer.
To achieve this feat, I let Solomon think that he came up with the Altair name, and for years he told his story at various trade-show conventions: He was inspired by his daughter, who was watching a Star Trek episode in which the starship Enterprise was on its way to a star named Altair. What a joke that was!
But the story was OK with me. As long as the Altair worked well enough to run the software I was hoping we'd find for it, and as long as other people made personal computers, the revolution was off to a nice start.
A Letter to Bill
The hardware was happening, but software would be a problem. I could see that there wasn't a soul in Albuquerque who could write good software, so I needed to look in faraway places. Ed Roberts wasn't much help in the matter. He was a hardware guy; he looked down on software as "trivial" and "nothing but variable hardware."
Without software, there was very little a user could do with the Altair other than flip toggle switches to create machine-language code that controlled the patterns of the little red front-panel lights. I need someone really smart who was willing to work his brains out for very little money.
I had no idea how to find this person, but as it turned out he found me. One day Ed told me that a junior engineer at Digital Equipment named Paul had called, claiming that his friend Bill—who was still a freshman at Harvard—could write software for the Altair. Ed wasn't convinced. He felt that these guys were too young to know what they were doing, so he turned them over me.
I called Paul and asked him, "Hey, can you guys write a Basic programming language for the Altair computer?" I suggested basing it on the version of Basic written by Professor John Kemery at Dartmouth.
"Sure I can," Paul replied. "But I've got to get Bill to go along with it. And his dad is going to be really pissed if he drops out of Harvard. Perhaps you should write him a letter."
And this how I came to write the famous letter (see The Letter) that convinced Bill Gates to drop out of Harvard, move to Albuquerque, write a Basic for the masses, start Microsoft and become the richest man in the world.
Bill and Paul called back the next day, just as Ed was telling me it might be better to equip the Altair with Cobol rather than Basic because "Cobol is the programming language for business." I was freaking out at that idea. Cobol was too complicated and was absolutely devoid of eloquence. If Ed got his way with Cobol, I thought, we could kiss the PC revolution good-bye.
Luckily, Bill and Paul said they would write Basic for the Altair and would consider moving to Albuquerque if Ed made them the right deal—including the right to license Basic to other computer makers. Paul would work for us as VP of software, and Bill would be a contract employee while he was setting up his own software company, which I suggested they call MicroSoft. They liked the name but later changed the capital "S" to lowercase, which in my mind was a sign that they had sold out to the establishment.
A couple of weeks after that call, Paul hopped on a plane and flew off to Albuquerque to show Ed and me the prototype version of Basic. The software was encoded on punch tape; Paul loaded it into the Altair through a teletype machine, which was then a popular type of terminal. Paul typed in a few commands, and the program responded by printing out the message "Hello."
The Revolution to Come
Bill and Paul, along with a third guy named Marty, moved into the Sundowner Motel across the street from MITS and began coding the first release version of Basic. When they weren't hard at work 15 hours a day, we hung out at a neighborhood restaurant eating endless bowls of killer green-chili stew and debating the future of the personal computer revolution.
The three of us saw eye-to-eye on the need for applications like word processors and file managers, but Paul and Bill didn't believe my predictions that the personal computer would evolve into a multimedia communications tool, that millions of PCs around the world would be connected to one another through a global network and that people would be sending each other video e-mail.
Truth was, both Bill and Paul were fairly wet behind the ears in those days. I was the one who got them to think on a grander scale.
Over the course of the next few years, I helped the two understand the need for a standardized PC operating system, and I explained to them the strategy of providing the applications as well.
Meanwhile, I was busy starting computer magazines and inventing things like spreadsheets and screen savers. But those are stories for another day.
The letter that convinced Bill Gates to drop out of Harvard, move to Albuquerque and become the richest man in the world.
MITS 2485 Lynn Ave. Albuquerque, N.M. 87106
Cambridge, Mass. 02138
February 15, 1975
You don't know me yet, but your friend Paul Allen tells me you are a hell of a poker player and also very smart at programming computers. Believe it or not, you might just be the dude I'm looking for.
My name's David Bunnell and I'm the vice president of business development at MITS here in Albuquerque. Paul says you saw the Altair on the cover of Popular Electronics and you flipped out. Paul also told me you're interesting in writing software for the Altair computer.
My friend and boss Ed Roberts and I like to tinker around with electronic stuff, and we built the Altair around this hot new microcomputer chip, the Intel 8080.
You see, the Altair is a fully functional, 8-bit programmable computer that has the potential to do just about anything you can imagine. And electronics hobbyists are lusting for our machines. We're so backlogged that some will do almost anything to get their hands on one: They send us cash in envelopes, and right now there's a guy camped out in his van awaiting delivery of his Altair.
Our biggest impediment to really big growth is the sad reality that right now you can program the Altair only through these front-panel toggle switches in machine language, which is a real pain. Unfortunately, all anyone can do is make these little red LEDs blink in psychedelic patterns. My hobbyist friends dig it, but I want to sell lots of these things and become a billionaire.
Ordinary people will never buy an Altair just to see its lights blink, Bill, and this is where I think you fit into the picture.
I'm not really into coding, even though I totally understand it. Paul tells me you know how to generate tight, tight code, which is really radical because the Altair has only 256 bytes of memory. Once memory-chip prices come down, it will have 65K bytes, but right now it is a bit skimpy in this department.
Ed and I have some pretty good ideas about how we can make this plan work, but of course the big question is whether you're willing to drop out of Harvard University and move to Albuquerque.
As I discussed with Paul, if you can copy Dartmouth Basic and make it work on the 8080 processor, then we can bundle it with the Altair and jointly license it to other personal computer makers. Your software and our hardware will sell like hotcakes to all the electronics buffs, who will in turn write programs that will make personal computers useful for ordinary people. We'll sell millions of these machines, all with your software.
Our goal should be to have a computer that runs your software on every office desktop and in every home—first in America and then the world. I'll make you filthy rich, Bill Gates, and that's a promise!!
All you've got to do is quit wasting your time playing poker, drop out of college and head on out to New Mexico. I know that Albuquerque isn't exactly glamorous, but the Mexican food is fantastic, and there are these flat desert areas where you can race cars and do wheelies, which I understand you like to do.
Please say YES, Bill. Paul will smooth things over with your folks up in Seattle. If you'd like, I'll even write a letter to your Dad explaining the whole situation and why you'll do well here.
Let me know soon, though—there's this Navy guy named Gary Kildall who also wants to write software for the Altair, and he's always calling Ed from San Diego or Honolulu. Ed favors military guys, so please get back to us as soon as you can.
Vice President of Advertising, Marketing, Technical Publications & New Product Development