You know how a TV scans one line at a time on your TV, from top to bottom? It takes about 65 microseconds (millionths of a second) to scan each line on a U.S. TV. Well, it turns out that about 40 of these microseconds are visible and the other 25 microseconds are not. During this 25-microsecorid time, the so- called refresh period, I inserted 16 unique addresses to the DRAM. (I got these addresses for free, using the counters of the terminal, which were generating video signals.)
I had selection chips that selected the address to come from the horizontal and vertical counter chips of the terminal during this period. Amazingly, it only took two of these selection chips and maybe another chip or two worth of logic to do the whole thing. So I actually stole some cycles away from the microprocessor to refresh the DRAM.
I would Ve had no idea how to get a DRAM chip, but luckily, right around this time someone at the club who worked at AMI offered some 4K-bit DRAM chips for sale at a reasonable price. This was before they were even on the market. I see now that someone must've ripped them off from AMI, but I didn't ask any questions.
I bought eight of them from the AMI guy for about $5 each and
modified my design. I added some wires to the memory connector on the Apple I board so it could accommodate either an SRAM or DRAM board. I plugged the new DRAM board in, and it worked the very first time.
• o •
I had been showing off this exciting design of mine to Steve Jobs. He'd gone with me to Homebrew a few times, helping me carry in my TV. He kept asking me if I could build a computer that could be used for time-sharing—like the minicomputer a local company called Call Computer used.
The year before, Steve and I had sold my ARPANET terminal to Call Computer in Mountain View, giving them the rights to build and sell it.
"Sure," I said. "Someday." It could be done, I thought, but it was ages off.
Then he asked if 1 could add a disk for storage someday. I said, again, "Sure. Someday." This all seemed a long way off.
Then, a few days after I got the AMI DRAMs working, Steve called me at work. He asked me if I'd considered using the Intel DRAMs instead of AMI's.
"Oh, Intel's are the best, but I could never afford them," I told him.
Steve said to give him a minute.
He made some calls and by some marketing miracle he was able to score some free DRAMs from Intel—unbelievable considering their price and rarity at the time. Steve is just that sort of person. I mean, he knew how to talk to a sales representative. I could never have done that; I was way too shy.
But he got me Intel DRAM chips. Once I had them, I redesigned around them. And I was so proud because my computer looked smaller yet. I had to add a couple more chips to my computer to make it work with the Intel DRAMs. But the Intel chips were physically so much smaller than the AMI chips.
I have to stop here and explain what the big deal about having a smaller-sized chip is. Remember when I said my goal since high school had always been to have the fewest chips? Well, that isn't the whole story. One time in high school, I was trying to get chips for a computer I'd designed. My dad drove me down to meet an engineer he knew at Fairchild Semiconductor, the company that invented the semiconductor. I told him I'd designed an existing minicomputer two ways. I found out that if I used chips by Signetics (a Fairchild competitor), the computer had fewer chips than if I used Fairchild chips.
The engineer asked me which Signetics chips I'd used.
I told him the make and model number.
He pointed out that the Signetics chips I'd used in the design were much larger in physical size, with many more pins and many more wires to connect, than the equivalent Fairchild chips. That added complexity.
I was stunned. Because he made me realize in an instant that the simpler computer design would really have fewer connections, not simply fewer chips. So my goal changed, from designing for fewer chips to trying to have the smallest board, in square inches, possible.
Usually fewer chips means fewer connections, but not always.
Back to the Intel DRAM design of the Apple I, switching from AMI to Intel DRAM memories meant I could reduce the total size of the board, even though I had to add a couple extra chips to do it.
And looking back, what a great, lucky decision it was to go with Intel's chips. Because that chip design eventually became the standard for all memory chips, even to this day.
• o •
By Thanksgiving of 1975, Steve had been to a few of the Homebrew meetings with me. And then he told me he'd noticed something: the people at Homebrew, he said, are taking the
schematics, but they don't have the time or ability to build the computer that's spelled out in the schematics.
He said, "Why don't we build and then sell the printed circuit boards to them?" That way, he said, people could solder all their chips to a printed circuit (PC) board and have a computer in days instead of weeks. Most of the hard work would already be done. His idea was for us to make these preprinted circuit boards for $20 and sell them for $40. People would think it was a great deal because they were getting chips almost free from their companies anyway.
Frankly, I couldn't see how we would earn our money back. I figured we'd have to invest about $1,000 to get a computer company to print the boards. To get that money back, we'd have to sell the board for $40 to fifty people. And I didn't think there were fifty people at Homebrew who'd buy the board. After all, there were only about five hundred members at this point, and most of them were Altair enthusiasts.
But Steve had a good argument. We were in his car and he said—and I can remember him saying this like it was yesterday: "Well, even if we lose our money, we'll have a company. For once in our lives, we'll have a company."
For once in our lives, we'd have a company. That convinced me. And I was excited to think about us like that. To be two best friends starting a company. Wow. I knew right then that I'd do it. How could I not?
Chapter 12
Our Very Own Company
To come up with the $1,000 we thought we'd need to build ready-made printed circuit boards, I sold my HP 65 calculator for $500. The guy who bought it only paid me half, though, and never paid me the rest. I didn't feel too bad because I knew HP's next-generation calculator, the HP 67, was coming out in a month and would cost me only $370 with the employee discount.
And Steve sold his VW van for another few hundred dollars. He figured he could ride around on his bicycle if he had to. That was it. We were in business.
Believe it or not, it was only a couple of weeks later when we came up with a name for the partnership. I remember I was driving Steve back from the airport along Highway 85. Steve was coming back from a visit to Oregon to a place he called an "apple orchard." It was actually some kind of commune.
Steve suggested a name—Apple Computer.
The first comment out of my mouth was, "What about Apple Records?" This was (and still is) the Beatles-owned record label.
We both tried to come up with technical-sounding names that were better, but we couldn't think of any good ones. Apple was so much better, better than any other name we could think of. Steve didn't think Apple Records would have a problem since it probably was a totally different business. I had no idea.
So Apple it was. Apple it had to be.
• o •
Really soon after that, we met with a friend of Steve's who worked at Atari. This guy said he'd be able to design the basic layout of my printed circuit board, based on my original design, for about $600. That was what we needed so we could take it into a manufacturing company that could mass-produce boards.
We also met with another guy from Atari, Ron Wayne, who Steve thought could be a partner. I remember meeting him for the first time and thinking, Wow, this guy is amazing. He could just sit at a typewriter and type out our whole legal partnership agreement like he's a lawyer. He wasn't a lawyer, but he knew all the legal words. He was a fast talker and he seemed so smart. He was one of those people who seemed to have a quick answer for everything. He seemed to know how to do all the things we didn't.
Ron ended up playing a huge role in those very early days at Apple—this was before we had funding, before we'd done much of anything. He was really the third partner, when I think of it. And he did a lot. He wrote and laid out the early operation manual. After all, he could type stuff. And he could draw. He was the one who did the etching of Newton under the Apple tree that was on the computer manual.
Underneath it was a line from a William Wordsworth poem describing Newton. It said: "A mind forever voyaging through strange seas of thought. . . alone."
Eventually Steve, Ron, and I figured out a partnership agreement that started Apple and included all three of us. Steve had 45 percent, I had 45 percent, and Ron got 10 percent. We both trusted him as someone who'd be able to resolve arguments. Ron started working on the paperwork.
• o •
Where Did That Weird Quote Come From?
I had to look this one up. It turns out it is from book 3 of
The
Prelude
by William Wordsworth. (A Mind Forever Voyaging is also the name of a video game from 1985. Who knew?) The lines in full read like this:
The antechapel where the statue stood
Of Newton with his prism and silent face,
The marble index of a mind for ever
Voyaging through strange seas of Thought, alone.
Before the partnership agreement was even inked, I realized something and told Steve. Because I worked at HP, I told him, everything I'd designed during the term of my employment contract belonged to HP.
Whether that upset Steve or not, I couldn't tell. But it didn't matter to me if he was upset about it. I believed it was my duty to tell HP about what I had designed while working for them. That was the right thing and the ethical thing. Plus, I really loved that company and I really did believe this was a product they should do. I knew that a guy named Miles Judd, three levels above me in the company structure, had managed an engineering group at an HP division in Colorado Springs that had developed a desktop computer.
It wasn't like ours at all—it was aimed at scientists and engineers and it was really expensive—but it was programmable in BASIC.
I told my boss, Pete Dickinson, that I had designed an inexpensive desktop computer that could sell for under $800 and would run BASIC. He agreed to set up a meeting so I could talk to Miles.
I remember going into the big conference room to meet Pete, his boss, Ed Heinsen, and Ed's boss, Miles. I made my presentation and showed them my design.
"Okay," Miles said after thinking about it for a couple of minutes. "There's a problem you'll have when you say you have output to a TV. What happens if it doesn't look right on every TV? I mean, is it an RCA TV, a Sears TV, or an HP product that's at fault?"
HP keeps a close eye on quality control, he told me. If HP couldn't control what TV the customer was using, how could it make sure the customer had a good experience? More to the point, the division didn't have the people or money to do a project like mine. So he turned it down.
I was disappointed, but I left it at that. Now I was free to enter into the Apple partnership with Steve and Ron. I kept my job, but after that I was officially moonlighting. Everybody I worked with knew about the computer board we were going to sell.
Over the next few months, Miles would keep coming up to me. He knew about BASIC-programmable computers because of his division out in Colorado, and even though they didn't want my design, he said he was intrigued by the idea of having a machine so cheap that anyone could own one and program it. He kept telling me he'd been losing sleep ever since he heard the idea.
But looking back, I see he was right. How could HP do it? It couldn't. This was nowhere near a complete and finished scientific engineer's product. Everybody saw that smaller, cheaper computers were going to be a coming thing, but HP couldn't justify it as a product. Not yet. Even if they had agreed, I see now that HP would've done it wrong anyway. I mean, when they finally did it in 1979, they did it wrong. That machine went nowhere.
A few weeks after the meeting, the PC board was finished and working. I was so proud of it. I was at HP showing it off to some engineers when the phone rang at the lab bench.
It was Steve.
"Are you sitting down?"
"No," I told him.
"Well, guess what? I've got a $50,000 order."
"What?"
Steve explained that a local computer store owner had seen me at Homebrew and wanted to buy one hundred computers from us. Fully built, for $500 each.
I was shocked, just completely shocked. Fifty thousand dollars was more than twice my annual salary. I never expected this.