iWoz (20 page)

there was holding up the magazine
Popular Electronics
, which had a picture of a computer on the front of it. It was called the Altair, from a New Mexico company named MITS. You bought the pieces and put them together and then you could have your own computer.

So it turned out all these people were really Altair enthusiasts, not TV terminal people like I thought. And they were throwing around words and terms I'd never heard—talking about microprocessor chips like the Intel 8080, the Intel 8008, the 4004,1 didn't even know what these things were. Like I said, I'd been designing calculators for the last three years, so I didn't have a clue.

I felt so out of it—like, No, no, I am not of this world. Under my breath, I am cussing Allen Baum. I don't belong here. And when they went around and everyone introduced themselves, I said, "I'm Steve Wozniak, I work at Hewlett-Packard on calculators and I designed a video terminal." I might have said some other tilings, but I was so nervous at public speaking that I couldn't even remember what I said afterward. After that, we all signed a sheet of paper where we were supposed to put clown our name and what interests and talents we were bringing to the group. (This piece of paper is public now; you might be able to find it online.) The thing I wrote on that paper was, "I have very little free time."

Isn't that funny? These days I'm so busy and people are constantly asking for my autograph and stuff, but back then I was also just as busy: always working on projects, engineering for work and then engineering at home. I don't feel like I've changed much since then, and I guess this proves it, sort of.

Well, anyway, I was scared and not feeling like I belonged, but one very lucky thing happened. A guy started passing out these data sheets—technical specifications—for a microprocessor called the 8008 from a company in Canada. (It was a close copy,

or clone, of Intel's 8008 microprocessor at the time.) I took it home, figuring, Well, at least I'll learn something.

• o •

That night, I checked out the microprocessor data sheet and I saw it had an instruction for adding a location in memory to the A register. I thought, Wait a minute. Then it had another instruction you could use for subtracting memory from the A register. Whoa. Well, maybe this doesn't mean anything to you, but I knew exactly what these instructions meant, and it was the most exciting thing to discover ever. Because I could see right away that these were exactly like the instructions I used to design and redesign on paper for all of those minicomputers back in high school and college. I realized that all those minicomputers I'd designed on paper were pretty much just like this one.

Only now all the CPU parts were on one chip, instead of a bunch of chips, and it was a microprocessor. And it had pins that came out, and all you had to do was use those pins to connect things to it, like memory chips.

Then I realized what the Altair was—that computer everyone was so excited about at the meeting. It was exactly like the Cream Soda Computer I'd designed five years before! Almost exactly. The difference was that the Altair had a microprocessor—a CPU on one chip—and mine had a CPU that was on several chips. The other difference was that someone was selling this one—for $379, as I recall. Other than that, there was pretty much no difference. And I designed the Cream Soda five years before I ever laid eyes on an Altair.

It was as if my whole life had been leading up to this point. I'd done my minicomputer redesigns, I'd done data on-screen with Pong and Breakout, and I'd already done a TV terminal. From the Cream Soda Computer and others, I knew how to connect memory and make a working system. I realized that all I needed was this Canadian processor or another processor like it and

some memory chips. Then I'd have the computer I'd always wanted!

Oh my god. I could build my own computer, a computer I could own and design to do any neat things I wanted to do with it for the rest of my life.

I didn't need to spend $400 to get an Altair—which really was just a glorified bunch of chips with a metal frame around it and some lights. That was the same as my take-home salary, I mean, come on. And to make the Altair do anything interesting, I'd have to spend way, way more than that. Probably hundreds, even thousands of dollars. And besides, I'd already been there with the Cream Soda Computer. I was bored with it then. You never go back. You go forward. And now, the Cream Soda Computer could be my jumping-off point.

No way was I going to do that. I decided then and there I had the opportunity to build the complete computer I'd always wanted. I just needed any microprocessor, and I could build an extremely small computer I could write programs on. Programs like games, and the simulation programs I wrote at work. The possibilities went on and on. And I wouldn't have to buy an Altair to do it. I would design it all by myself.

That night, the night of that first meeting, this whole vision of a kind of personal computer just popped into my head. All at once. Just like that.

• o •

And it was that very night that I started to sketch out on paper what would later come to be known as the Apple I. It was a quick project, in retrospect. Designing it on paper took a few hours, though it took a few months longer to get the parts and study their data sheets.

I did this project for a lot of reasons. For one thing, it was a project to show the people at Homebrew that it was possible to build a very affordable computer—a real computer you could

program for the price of the Altair—with just a few chips. In that sense, it was a great way to show off my real talent, my talent of coming up with clever designs, designs that were efficient and affordable. By that I mean designs that would use the fewest components possible.

I also designed the Apple I because I wanted to give it away for free to other people. I gave out schematics for building my computer at the next meeting I attended.

This was my way of socializing and getting recognized. I had to build something to show other people. And I wanted the engineers at Homebrew to build computers for themselves, not just assemble glorified processors like the Altair. I wanted them to know they didn't have to depend on an Altair, which had these hard-to-under- stand lights and switches. Every computer up to this time looked like an airplane cockpit, like the Cream Soda Computer, with switches and lights you had to manipulate and read.

Instead they could do something that worked with a TV and a real keyboard, sort of like a typewriter. A computer like I could imagine.

As I told you before, I had already built a terminal that let you type regular words and sentences to a computer far away, and that computer could send words back to the TV. I just decided to add the computer—my microprocessor with memory—into the same case as that terminal I'd already built.

Why not make the faraway computer this little microprocessor that's right there in the box?

I realized that since you already had a keyboard, you didn't need a front panel. You could type things in and see things on-screen. Because you have the computer, the screen, and the keyboard, too.

So people now say this was a far-out idea—to combine my terminal with a microprocessor—and I guess it would be for other people. But for me, it was the next logical step.

That first Apple computer I designed—even though I hadn't named it an Apple or anything else yet—well, that was just when everything fell into place. And I will tell you one thing. Before the Apple I, all computers had hard-to-read front panels and no screens and keyboards. After Apple I, they all did.

• o •

Let me tell you a little about that first computer—what is now called the Apple I—and how I designed it.

First, I started sketching out how I thought it would work on paper. This is the same way I used to design minicomputers on paper in high school and college, though of course they never got built. And the first thing was I had to decide what CPU I would use. I found out that the CPU of the Altair—the Intel 8080—cost almost more than my monthly rent. And a regular person couldn't purchase it in small or single-unit quantities anyway. You had to be a real company and probably fill out all kinds of credit forms for that.

Luckily, though, I'd been talking to my cubicle mates at HP about the Homebrew Club and what I was planning, and Myron Tuttle had an idea. (You remember him: the guy whose plane almost crashed when I was in it.) He told me there was a deal you could get from Motorola if you were an HP employee. He told me that for about $40, I could buy a Motorola 6800 microprocessor and a couple of other chips. I thought, Oh man, that's cheap. So very quickly I knew exactly what processor I would have.

Another thing that happened really early on was I realized— and it was an important realization—that our HP calculators were computers in a real sense. They were as real as the Altair or the Cream Soda Computer or anything else. I mean, a calculator had a processor and memory. But it had something else, too, a feature computers didn't have at the time. When you turned a calculator on, it was ready to go: it had a program in it that started up and then it was ready for you to hit a number. So it booted up automatically and just sat there, waiting for you to tell it to do some thing. Say you hit a "5." The processor in the calculator can see that a button is pushed, and it says, Is that a 1? No. A 2? No. A 3, 4 ... it's a 5. And it displays a 5. The program in a calculator that did that was on three little ROM (read-only memory) chips— chips that hold their information even if you turn the power off.

So I knew I would have to get a ROM chip and build the same kind of program, a program that would let the computer turn on automatically. (An Altair or even my Cream Soda Computer didn't do anything for about half an hour after you set switches so you could put a program in.) With the Apple I, I wanted to make the job of having a program go into memory easier. This meant I needed to write one small program which would run as soon as you turned your computer on. The program would tell the computer how to read the keyboard. It would let you enter data into memory, see what data was in memory, and make the processor run a program at a specific point in memory.

What took about half an hour to load up a program on the Altair, took less than a minute using a keyboard on the Apple I.

What Is ROM?

Read-only memory (ROM) is a term you'll hear a lot in this book. A ROM chip can only be programmed once and keeps its information even if the power is turned off. A ROM chip typically holds programs that are important for a computer to remember.

Like what to do when you turn it on, what to display how to

recognize connected devices like keyboards, printers, or monitors. In my Apple I design, I got the idea from the HP calculators (which used two ROM chips) to include ROMs. Then I could write a "monitor" program so the computer could keep track of what keys were being pressed, and so on.

If you wanted to see what was in memory on an Altair, it might take you half an hour of looking at little lights. But on the Apple I, it took all of a second to look at it on your TV screen.

I ended up calling my little program a "monitor" program since that program's main job was going to be to monitor, or watch, what you typed on the keyboard. This was a stepping point—the whole purpose of my computer, after all, was to be able to write programs. Specifically, I wanted it to run FORTRAN, a popular language at the time.

So the idea in my head involved a small program in read-only memory (ROM) instead of a computer front panel of lights and switches. You can input data with a real keyboard and look at your results on a real screen. I could get rid of that front panel entirely, the one that made a computer look like what you'd see in an airplane cockpit.

Every computer before the Apple I had that front panel of switches and lights. Every computer since has had a keyboard and a screen. That's how huge my idea turned out.

• o •

My style with projects has always been to spend a lot of time getting ready to build it. Now that I saw my own computer could be a reality, I started collecting information on all the components and chips that might apply to a computer design.

I would drive to work in the morning—sometimes as early as 6:30 a.m.—and there, alone in the early morning, I would quickly read over engineering magazines and chip manuals. I'd study the specifications and timing diagrams of the chips I was interested in, like the $40 Motorola 6800 Myron had told me about. All the while, I'd be preparing the design in my head.

The Motorola 6800 had forty pins—connectors—and I had to know precisely how each one of those forty pins worked. Because I was only doing this part-time, this was a long, slow process. And several weeks passed without any actual construc

tion happening. Finally I came in one night to draw the design on paper. I had sketched it crudely before. But that night I came in and drew it carefully on my drafting board at Hewlett-Packard.

It was a small step from there to a completely built computer. I just needed the parts.

• o •

I started noticing articles saying that a new, superior-sounding microprocessor was going to be introduced soon at a show, WESCON, in San Francisco. It especially caught my attention that this new microprocessor—the 6502 from MOS Technologies in Pennsylvania—would be pin-for-pin compatible with, electrically the same as, the Motorola 6800 I had drafted my design around. That meant I could just pop it in without any redesigning at all.