This is a transcript of an audio interview. This transcript may contain errors - if you're using this material for research, etc. please verify with the original recorded interview.
Source: ANTIC: The Atari 8-Bit Podcast
Interviewer: Kevin Savetz
KS: I’m Kevin Savetz, and this is an interview episode of Antic, the Atari 8-bit podcast. Bill Mensch is co-creator of the 6502 chip, the microprocessor that’s the heart of the Atari 8-bit computers, the Apple II, the Commodore 64, and many other classic machines. This interview occurred August 6, 2015
KS: You can start by telling me, what’s it take to develop a processor, and what did you have in mind when you developed the 6502?
BM: Yeah so what happens is… a little bit of background. Not trying to take up too much of your time in the interview process here butm actually I’m a hardware guy, but actually I started out more or less in software. So I’m actually a software guy converted to a hardware guy. The software that I did was at the University of Arizona. At the University of Arizona, for my senior year, my second year because I only spent two years working on my BS. For my senior year, I supported myself with a half-time job working for a professor that came back from Princeton… Princeton. At Princeton his sabbatical was working with the Tokamak fusion project they were working on, and when he came back to the University of Arizona, he wanted to simulate the hardware that he was working on at Princeton. So he’s a nuclear physicist professor… so I was taking a full load of engineering classes, so I could graduate in two years, plus I was taking six credit hours of Newton graduate-level nuclear physics… so that that was required for my half-time minimum wage job, simulating fusion. So that got me into the software world, and as a result when I did get my job my first job at Motorola, I simulated everything. So that means that I simulated the integrated circuits, so that I knew the transistor functions were working correctly, and when you put two or more transistors together with resistors, all of the equations for the transistor operation, due to voltages and currents are all in the simulation of the transistor. So I was simulating the transistors, so that I would know that when the chip came out, the chip would work. And so I also simulated the process that manufactured the transistor. So I knew everything about the way the physics of semiconductors worked, with voltages and currents and resistive characteristics, and so I built up my understanding through my software, simulating everything about an integrated circuit. So once you do that, you realize that actually a microprocessor is a small computer. So a microcomputer is just a name for a small computer, and so if you’re working, as I was, the language that are programmed and was Fortran is still used by astrophysicists, because it is the language that works most efficiently for formulas. That’s what it’s called, “formula translation”. I think is what Fortran. Fortran was the first high-level language also. So now, that simulation led me to understanding, like at the University of Arizona, working for this nuclear physicist professor, for fusion, when I.. he wanted me to get my PhD in nuclear physics, and when I asked him… and also, back up a minute… I took over a PhD dissertation, so this guy graduated with a PhD in nuclear physics, and so nobody else was available in electrical engineering, which is what he wanted, so I was the only available student that filled the criteria that he was looking for, and nobody else really wanted to work, and take all this extra effort. But this was a choice that I made, because of the situation I was going through. And so as a result then, I would simulate this fusion… this fusion simulation, and I had the computer for two hours in the middle of the night. Let’s say from 2 o’clock in the morning until 4 o’clock in the morning. So when I asked the professor. And that’s because it was… It cost $300 an hour for the entire computers’ facility, using every peripheral. Every bit of the power. No other program could run when I was running that computer. So essentially had a personal computer, at the University of Arizona in 1970-‘71. Now how many times would have to go to a computer building with a card deck, in the middle of the night, before you’d say “well you know what, if I had my own computer I could do this 24x7, but the other thing is that I wouldn’t have to get up in the middle of the night.” And so that’s just some basic thinking that you go through. So what happened was, I asked them how soon they were going to cap.. you know, harness fusion energy as an energy source, and he said their goal was by the year 2025, which is in about another 10 years. I said will wait a minute that takes longer than it takes to build a house. That’s a long time. Wow! So let’s see. 1970.. 2025. Looks like about what about 55 years? So I said, well no, I really want to get a job, so not gonna take you up on your offer for me to get a PhD here. So that’s when I realized that having your own compute,r would probably benefit, because not only benefit the person having their own computer, the maybe we could collapse the 55 years down to maybe something more reasonable like, 20 years. So that was my thinking. So in 1973, after having designed not only a computer chip for Olivetti, at Motorola, and the 6800, and the peripherals, in the middle of the design of the PIA, which is an I/O chip, I suggested to check Chuck Peddle, that we could do a home computer. And this would’ve been around December 1973. So, I wasn’t planning to do a home computer. I was suggesting a home computer would be what you could do when you combine a microprocessor with I/O and memory, both data memory and program memory. Which is about same time that Hewlett-Packard 35 or 45, or whatever the number was, having a fixed function of mathematical functions, like add, subtract, multiply ,divide, and square root and some other things that the HP calculator did. So all I was doing the saying a general-purpose microprocessor, with general-purpose I/O, and the ability to load programs in it, would be that personal, I didn’t call personal, that home computer, that accountants could use at the same time, I happened to be installing sprinkler systems in Arizona. If you don’t sprinkle the grass it dies, so you need sprinkler systems. So that sprinkler system, each one of the valves is costing like I don’t know $60, or something like that, so if you just said you had a few valves and you had a controller, pretty soon you’d say I could replace everything but the valves, with a home computer. And so that’s what I was thinking in 1973, because I had patents on all of the integrated circuits associated with the 65… 6800. So Chuck came to me and said “Bill, do you want to go to Austin, Texas, that’s where they’re going to move us?” I said “no, I’ve driven through Texas a few times, coming back and forth from Pennsylvania to Arizona, and the only thing good about Texas is leaving Texas.” So for me to stay in the heart of Texas, with Austin Texas is where they moved the microprocessor group to, I said I’d rather have a different choice. And so Chuck came to me, because he knew I knew the whole thing, and under the process, I knew all the semiconductor engineering of it, and as a result having proven myself, and working on him on the PIA, and the microprocessor. So he said “well how about if we build a team to design a microprocessor, and will figure out where it is, because these guys at Motorola aren’t going to do the microprocessor that we need to do, and that is a low-end microprocessor to compete with the Intel 4040, which sold for about $29.” I said “right, that’s the place to go. Because that’s going to be embedded in everything that needs a microprocessor.” So what happens is, he comes to me, we meet with a guy named Paul Barrosa, who was the head of central research at Motorola, thinking that Paul Barrosa would help us figure out how to start a company, or find a company, that wanted to acquire a team that wants to do a great microprocessor. Now we already… I had already designed two microprocessors, and I, at that point, understood what it took to build a microprocessor with an advanced manufacturing process called depletion mode load N-MOS, N-channel MOS technology. Motorola didn’t have such a process. Intel didn’t have such a process. But as a semiconductor engineer, we knew that depletion loads were going to be better than the enhancement mode loads that we had to deal with, and we were the only company, Motorola was the only company, to do a 5-volt microprocessor. Intel had +12,-12 and 5 volts. So that’s three power supplies to do a microprocessor. Motorola had a single 5-volt supply on the 6800. And as a result, the technology to build a 5-volt-only microprocessor, using enhancement-only devices, requires a lot more transistors, and a lot more complication, but we got all of our circuits to work, we knew how to do it, but I knew that the depletion load process, would be the best process for the microprocessor technology. Now, how do I know that? I designed the microprocessor… all the microprocessors were built on the same manufacturing line, I designed the manufacturing control monitor that was stepped into every wafer of the 6800. I don’t know when they retired my process control monitor, but I had both enhanced devices for loads, and also depletion, in the same orientation on that monitor. All they had to do was implant it, and Motorola could add the process that we had at MOS technology. They did not want to do that until basically I was working out the door. So when I got to MOS technology I said “that’s the process we need”. We get that process, we have less transistors. It’ll be higher speed. I will be lower power, and everything will work better. And that’s what they developed. So in one year, we created a brand-new process that Motorola and Intel did not have. They did not have at MOS technology. Theirs was all P-MOS. So we’re working with N-MOS and depletion loads, and when the chip came out, Rod Orgell was a logistics… he was a logic engineer, but he also knew semi-conductors. So when Chuck came to me to build a team, I said I need Rod Orgell. I want Ray Hurd to do the process control monitor. I went to school… Ray and I went to school at the University of Arizona. I actually hired my study partner. When I went to interview students at the U of A, for Motorola… after one-year on the job, they sent me to U of A to interview students. Then after I interviewed students, gave them the results of my interviews, they came back to me and said “we need one more engineer out of U of A, do you know somebody?” I said “yeah, my study partner. He’s studying on his MBA right now, I’ll ask him if he wants to work here.” He got the job; we worked together. So what happens is, I say to Chuck, “we need Rod Orgell, we need Ray Hurd, we need Mike Jaynes, and we need Harry Baucum.” That’s the semiconductor team that made the 6502 happen. Mike Jaynes and Harry Baucum were mask-designers. We get one more mask-designer after appearing at MOS Technology. And then Chuck got the production manager from MOS Technology, his name is Terry Holt. Then he also brought with him an applications engineer, from Motorola, his name is Will Mathis. Will Mathis, Rod Orgell, myself and Chuck Peddle, are the four inventors of the 6502 microprocessor. However, I’m the one that designed the decimal-correct circuitry that is… that’s what we had patented. Now going back to something… Chuck is a systems guy. Chuck understood a small system requirement. Chuck built the first point-of-sale terminal for Singer, I think back in New Jersey. So he knew small systems, but Chuck worked at General Electric’s mainframe computer facility in Phoenix, and that’s how the connection between MOS Technology and the Motorola team was… John Pevanin and was also at the GE mainframe, and he started MOS Technology with two other guys John McLaughlin and Mark Jaffe. So now, the connection was that John Pevanin, knew Chuck Peddle, and knew what was going on back in the GE days, and so that’s how we got connected, and so then I headed-up the semiconductor engineering of the 6502. Now there’s more to be talked about the 6502. The Atari uses the 6507 in the Atari games, and the 6532. Ray Hurt was the project engineer of the 6532, but I had already designed the 6530, that was used in the Kim system. So Ray Hurt and I have a patent on the 6532, and the combination of 6507, and the 6532 were the core of the Atari game systems.
KS: The VCS. The Atari 2600 game system.
BM: Yes. So then there’s another microprocessor, I don’t know if, it’s which number the Atari uses the 6502 or not, in the Atari computers… the 600, or the whatever it was, 800. But going back to that, the difference between the 6507 and the 6501. Actually the difference between the 6501 and 6502 was about a dozen transistors, that I suggested would be put on the 6501 to make it a 6502. The reason why suggested that, is Motorola on the 6800, and also with Intel in the 8080, they had a clock-generator chip that had to meet the characteristics for a non-overlapping clock. Now as a semiconductor engineer, and probably this is way out of your league of understanding, or at least…
KS: We’ve been out of my league of understanding for quite a while but keep going !
BM: Well, this is the truth… this is the fact. So will get to the right level, for where you’re going, but you are asking me about the 6502, and most of this is not known, or if it’s known, is not known in a coherent way by those experienced in the semiconductor-engineering world. So what happened was, Chuck is off on a marketing thing, while were designing the 6501, he comes back and I said “how about if we just put the clock-generator right on the chip,” and he said “what’s that gonna mean?” I said “well it’s not really mean anything, from a manufacturing cost standpoint. The chip is it going to be any different inside. I can hide the transistors underneath the metal, and so it’s not a bigger chip, there’s nothing… it’s not gonna cost anymore.” I just put the clock-generator on it, and the reason why, is because an external clock, if it’s manufactured on a different, which it would be, on a different processor, on a different wafer, the characteristics of that clock-generator are different. Like for instance, it could be a slow clock-generator, with a fast chip, and as a result these clocks are going to be overlapping, and it’s going to cause problems. So Motorola was selling the $69 clock-generator to go with a $375 6800. I said “we’ll throw a clock-generator on it, and we’ll make it an in-metal option. If you want a great. If you don’t want it, great.” So that’s the difference between a 6501 and a 6502. The reason why I had been… I had a bet with Rod Origall, and Rod Origall said the 6501 would outsell the 6502. And I said “well why do you think that?” He said “because of the marketing of Motorola. They’re going to be selling it, and when they plug in a 6800 with our PIA, they’re going to build it all around the 6800. We’ll just take the 6800 out, plug in a 6501, we’re good-to-go. All the sales and efforts by Motorola, we’ll just cannibalize it. We’ll get the business, because we’ll have a lot smaller, lower-cost chip than the 6800. The user isgoing to go with us, because we have a lower-cost, much lower-cost microprocessor.” I said “okay Rod, while we’re waiting for the chips to be actually manufactured,” it took about two-weeks, I said “Rod I’ll bet you the 6502 outsells the 6501. He said “No way. There’s no way.” Who’s going to be selling the 6502? I mean, okay we have a few guys like Will Mathis, and maybe Chuck going out beating-the-bushes, but don’t have the team that Motorola has selling the 6800, and if we have something plug-compatible, they’ll buy ours. I said “okay Rod, the bet’s on. I’ll bet you.” It could’ve been a six-pack, it could’ve been just a bet. Gentlemans’ bet-thing. I’ll bet. So when Motorola sued us, because we violated their patent, because we were actually doing things that already had patents on them. So we had patents on these things, Motorola comes along says “you can’t have a chip that plugs into a 6800 socket, so you have to stop selling the 6501.” And that day who, wins the bet right? So wait Rod, you just lost your bet. So that’s how became a 6502 engineer. And that’s the story. That’s my story, and I’m sticking to it Kevin.
KS: Nice. Nice. Were you surprised that it ended up being such a staple of all the home… many of the home computers over the next few years?
BM: No. We knew… Rod and I knew. Because we had designed two other microprocessors. We knew what Intel was doing. We knew what Motorola was doing. And really there was nobody else that counted. Look at it. Rockwell had their PPS4 and PPS8, but Rockwell ended up licensing ours, because they knew, at that point they designed their own microprocessor, they knew which was the best microprocessor. Now when you look at it, we knew what would be an awesome microprocessor, and most people do not know what I mean, because I’m the only one that’s calling it an addressable-register architecture. So, the programmers that program the Apples, Ataris, Commodores and Nintendos, they really didn’t know in a… let’s say in a conversation, why the 6502 was so powerful. And I don’t think the world really knows it, although we have benchmarks on our website, comparing the 6502, to the 65816 to 68000 and to the 8086. You can get online and see those benchmarks. I did not write the benchmarks. It’s an instruction-level benchmark.
KS: So tell me what makes it so powerful.
BM: Well, think about an eight… this is an 8-bit microprocessor, but it can compete very well with a 16-bit, and actually very well with a 32-bit processor. But what makes it so powerful, is that in one 8-bit quantity, you get an instruction. You don’t need any other information. Like for instance, increment the accumulator. Increment the index registers. You can do that with one 8-bit quantity. You cannot do that with the ARM processor. You cannot do that with some of the other processors out there. The 8086 can’t do that. So what happens here, is if you say everything you need is in 8-bit quantity, for 256 different instructions. So if you look at the opcode table of the 6502, you’ll see there’s a lot of “not used” opcodes. And the reason why some people call them… what’s the thing is... unusual operations, “illegal opcodes.” The reason for illegal opcodes, is we didn’t care what the opcodes did, we only cared what the ones that we wanted, would do. So if you put in a non-defined opcode, that some people have found, we never looked at it. We have… I have never specifically looked at, what the other opcodes would do on the 6502; the original. Because it wasn’t part of the design. If the opcode that you want works, that’s all that was required. So if somebody tested some of these other weird opcodes, you go “oh, so it decrements both the index-register and also the accumulator at the same time or whatever. Well, that’s interesting.” We didn’t know what it did, because we didn’t care about what it did.
KS: So that’s just pure luck of the design, what any illegal opcode would happen to do?
BM: Well, we never thought of them as illegal opcodes. They were called unused opcodes in the way we thought of it. So we weren’t using those opcodes. So they weren’t illegal. You can do whatever you want with them, but we weren’t going to tell you what they did, because we didn’t care what they did. So it’s not illegal. You’re just using it because you want to use it. So those were fun facts. What does this other opcode do? But look at the 65C02. Because I knew people were curious about those things. Guess what all the unused opcodes do. Do you know?
KS: I have no idea.
BM: They’re “no-operation.” Since they don’t affect any register. So that could have been disappointing to somebody that wanted to something weird with an unused opcode. So I give you the cycle-count for each of the unused opcodes in my data-sheet. So now you can say “oh, I have a no-op that uses two cycles” or whatever the no-op does, I think it’s two. No-operation, two cycles later, you didn’t do anything. Well all of the unused opcodes now have different cycle counts. Some of them are common, because of the way you decode the logic for the instruction. So actually there are no unused opcodes on the 816. So the one opcode that I reserved for, called WDM, that one is a no-operation, that one was meant to be a segue to another set of opcodes that I never did.
KS: Like bank-switching in, like we’re going to the second-set of opcodes, that sort of thing?
BM: Yeah. Like for instance, if you wanted to have a set of floating-point… Let’s say you wanted to have floating-point operations on the microprocessor. Then I could go… the first byte would be WDM, the second byte would be the floating-point operation.
KS: Why did you never do them?
BM: Because, you know, think about it. If you add all these other things in it, what are you getting? A speed-up, right? I already was doing what I wanted to do, and my microprocessor was outperforming the 68000 and the 8086, but when you trip over that boundary, to like the 8086… Look at what’s in the 8086 compared to the 816, look at what’s in the 68000, compared to the 816, and you’ll see that what you’re getting is a whole lot more complication, and that’s why, when my nephew worked for me to get his electrical engineering degree from ASU, both my nephews Rod Bearish, who I was just talking to… he’s a software engineer, and Andrew Hall, worked with, for IBM… Intel, after he left my company, he worked with them in Chandler, and then he was moved to Portland Oregon for designing the Pentium Pro microprocessor… and that Pentium Pro ran at 200 MHz at the time. Andrew worked with about 500 other engineers. So when you trip over…. going from my 816, to something competitive, you need about 1,000 engineers. And so that’s why I never did it. Now did I specify it? Can you find a datasheet for the 65c832? You can. Do we have a product name to the 65c832pxb, for programmable accelerator board, we do, we have one of those. We sell those. So the thing is, is that when you go from the 816, you have to add about 500 engineers. And the reason why, is because you have to compete with Intel, AMD, and at the time IBM. MIPS, you have to go to the Sparc, which came out of Stanford I think, no I don’t know which one came out of Stanford. I think MIPS came out of Stanford, and Sparc came out of Berkeley or something like that, so had the RISC-processors, you had Cambridge behind the ARM processors. And when you do that you have just elevated yourself to a whole new ecosystem, and I was happy with what I did with the ‘02 and 816. So every time I look at a 32-bit processor, I end up with “why not recommend the ARM processor?” Those guys came to me in 1983, to design the ARM for them, and I said “no, you’re asking me to do something that’s not compatible with the ‘02 and the 816.” They wanted to me to stop basically the 816, because the British broadcasting Company, British Empires Acorn, is where ARM came from. It was called Acorn RISC-Machine. And it was Advanced RISC Machine, now it’s just ARM, because it was never really a RISC-machine, but it followed the RISC-concept. And so therefore ARM doesn’t want to be known as a RISC-processor, and it’s because they really aren’t. However they have a lot of the characteristics. So now let’s go back to something. If you go back to the addressable-register architecture, you understand, we’ve had floating-point libraries for 25 years, 30 years. A floating-point operation is possible because of the way the 6502 instruction set works. So you can do a double-precision floating point number in software. That’s what we had. That’s the floating-point library. So if you want to add in a floating-point processor, like they did in the 486, when they went from 386 to 486, that’s when I bought for $125,000, I switched over from my Talma GDS2 system, to a network of Micron 486 processors, because I knew once they put the floating-point processor in there, then the software would use that, so all of my design tools would run on that processor. I switched from a Data General Eclipse to a network of, I think it was 15 or so Micron computers, we had to wait for the 486 to come off the production facility, before we could take delivery on, I think they were about $7000 per node, per PC. And so when you look at that the addressable-register architecture, if you can do all of these complex operations, like for instance encryption, you can do that with my processor. So the only thing the only thing you’re doing when you’re getting a 2 GHz quad-core Intel i7, all you did was speed up your software. But I can do this problem set with my 816. So then you go oh yeah, but speed is a big deal. I go “yeah, speed is a big deal, but look what it takes to get the speed there for a quad-core i7,” or however many cores they have now, I don’t even know how many, but the point is, it’s going to take manufacturing organization like Intel has, and you’re going to have to drive that down to about 20 nm technology to make sense out of that, and then as an end result, I’d recommend you buy it! I don’t have to design it. I really designed the ones I wanted, back in the 80s, about 30 years ago.
KS: So when you are designing that with your little team of four, six people, you said that you did simulation in software beforehand, but then when it came to actually designing the chip, are you actually sitting down at the drafting board with rubolyth and drawing things out... on plastic and paper?
BM: Back in that day we used mylar, because you could erase mylar without ripping the paper. So you use plastic pens… plastic lead, and so what you did then, was put the schematic… what we did… which was unique, not totally unique I guess, I don’t know if it was done by anybody else. it could have been. We do the whole chip on one piece of paper. And it could be a big piece of paper. So when you do it that way, it becomes the layout plan for the mask design. You follow that?
BM: So if you have a data-buffer here, on both the schematic and on the layout. If you have an address-bus across the chip, you have it right there. Now the way that Xerox was doing it, and I taught Xerox how to design microprocessors, after I started my company, and after Commodore and I parted ways, my first contract, after Commodore, was Xerox. Xerox in Southern California, in El Segundo, and Xerox PARC, in Northern California, were Jobs got his GUI. So Jobs is going into Northern California to Xerox, to take things, I’m going into Southern California version of Xerox, where they were doing advanced technology, I was teaching them how to design microprocessors. So I did that on the way to working with GTE, and licensing GTE, and Rockwell, and Centertech, I licensed all three of them on my ‘cO2. I licensed GTE on my 816, and some of the peripheral chips like the 22 chip and other things. I’ve licensed over 60 companies. So what happens here is, when you look at this from a semiconductor engineering standpoint, you know you have a basic building-block for anything that wants to have embedded intelligence. So you’re putting things into… software is what takes the microprocessor hardware and I/O and memory, and makes it into something that solves problems. So when I got the world’s most documented microprocessor, the world’s, probably the most famous microprocessor, I’m saying “why would I change from that and try to build, you know, a business out of something like Chuck Peddle did with the Commodore PET, or the Commodore 64.” Chuck didn’t design the Commodore 64, I don’t think, he didn’t have much at all to do with it, but that ended up being something that someone else needs to do. I’m an empowerment technology. I want to empower people to do their idea. That’s what I did for Chuck, that’s why he came to me. He knew I could build the chip that would meet the requirement that he thought he could use for his systems, but then he got lost. He went to the Intel 8086, at his Cirrus computer I think it was, and then between the Bill Gates, and Intel, and IBM, they just knocked Chuck right out of the market, because he was in a position to lead. And you have to lead and so I was leading. I have been leading for the last 37 years with the 6502 technology. Because I knew what it was. Rod Origell and I knew it. Rod says to me, “Bill, you know if this microprocessor works as good as it will you know we can go to the happy hunting ground.” That means that he can die, knowing that he did something special, with his career. Now Rod has passed away, others have passed away that were associated with this. And that’s what we knew. We knew we had the best. Because, one-byte for certain instructions, two-bytes for zero page, three-bytes for absolute addressing, and all I did with the 816 is add a third-byte to the address 24-bit addressing. You look at the time, 8086 had segments of I think it could handle 4 gigabytes… or 4 megabytes or something like that. And mine can handle 16 megabytes. So already had a bigger linear space than Intel had. Motorola had a bigger space, where Motorola screwed up, with the 68000, is instead of calling a 32-bit processor, they called it a 16-bit processor, because only had a 24 bit address bus, the same as mine. That was a mistake in marketing by Motorola. That should have always been a 32-bit processor because it hads 32-bit registers. The 8086 had 16-bit registers. The 8086 was really a 16-bit processor. The 68000 was always a 32-bit processor. And mine has always been an 8-bit processor with a 16-bit switch, for internal 16-bit registers. So someone that knows the hardware, knows that Motorola screwed up with marketing, that Intel out-marketed them, Intel always out-marketed them. Motorola went to the 8-bit or the microcontroller business, Motorola basically got out of the general-purpose microprocessor business. Intel took it over. AMD is just proving that they don’t have a monopoly. And there you go. So ARM came along, and ARM followed my business model. And that is, you have to survive on the design, on your design expertise. That’s what ARM took away that fateful day in about November of 1983. And when people say, “well what do you think about their stealing your ideas?” I said “they didn’t steal my ideas. I gave them ideas.” Because I already went to Gold St., Prudential Base wanted to take me public, and I had already decided I wanted to do it my way. And I don’t need any stockholders, I don’t need any…There’s never been another Board of Directors. It’s always been… I’ve been the president, CEO, chairman of the board, and the board of directors. So I never get into arguments. I just don’t have any negative input. And if I screw up… And if we were successful it was a team effort. If we screw-up, it was my fault.
KS: So where are you headed with the 6502? I know you have some hobbyist-oriented boards. Is that getting traction? Tell me about that.
BM: So the Xxcelr8r boards. So Xxcelr8r. Where did we come up with that name. Xxcelr8r uses the two x’s in the 65XX. But when you combine… if you call it 65Xxcelr8r, which we do on some of our packaging, I’m looking at it right now, we have 65xxcelr8r. What we came up with on one interesting Saturday, which is unusual for anyone to work except for me on a Saturday, although my guys work because they love what they do. Dave Kramer and I were talking on a Saturday, and I said “you know what, if you excel, that’s EXCEL, and you accelerate like in a car ACCEL, I said if you combine the two you have XX because you use XX and set of AC and EX, and use an eight because we love 8-bit, and you drop out a few other letters you have Xxcelr8r.” He said “that’s the one, you don’t have to think of another thing, that works.” So that’s where we got Xxcelr8r. So what are we accelerating? We are accelerating the understanding and application of embedded intelligent technology. So our branding, if you look at our board, had a big old X therefore Xxcelr8r, but what we have in the corner of the X, is S for sense, P for process, C for communicate, and A for actuate. That’s what embedded intelligence does. You sense your world; you process the information from the sensing of the world; you may communicate it to someone, like your wife or someone else; and then you’ll do something. And that’s what every piece of embedded intelligence in the universe does. So that’s what we’re teaching. The embedded intelligence. So have EIT in the middle of the X, and that’s what we’re trying to communicate. What we’re trying to do is enable others to understand embedded intelligence is what created the universe, and it’s created the matter, I’m saying something decided to matter, and going from energy to matter. So matter is created in the center of our stars all over the universe. That’s the combination of gravity, electric fields, and magnetic field, and somehow that decided to matter. And I don’t know how that decided to matter, or what was the trigger, except when I use my mind experiment, I decided I was so bored about sitting around statically as energy forever, and then I decided to wiggle, along with you and others, we decided to make matter. And why did we decide to matter. We decided to matter to understand ourselves, from our own senses. So this is all wrapped into the Xxcelr8r concept. And that’s the philosophy of embedded intelligence that you can found find on my foundation, the Mensch Foundation which is really a short form of the Bill and Diane Mitch Foundation for Embedded Intelligence Education. So now these boards have the basic general-purpose technology that’s been used forever with microprocessors, so if you look at the 6502 on the 6502 board, I have my 21 PIA that I invented, and the 22 that I invented, and then we have RAM, and flash memory, so you want to change memory for program space. Those are the basic building blocks of every microcomputer system in the world. Everything from Intel, everything from AMD, everything from MIPS, everything from Stanford and their technology and RISC, IBM. All the same components, they just run faster, they use up more power, they do interesting things, that we love. I’m talking on a very interesting thing, I don’t know if it’s quad-core ARM in this Moto X, or what it is, but that’s what I’m talking to you. Android operating system running on ARM processors. So what I’m saying is, our boards don’t compete with Arduino, don’t compete with raspberry pi, don’t compete with Beagle board, don’t compete with any other single-board computer known to mankind. They’re ours. They use our chips, and when… we’re working right now David Gray is putting a regulator on it, so our 5-volt SXB’s, work at 3.3 V so you can connect them up to what I call a programmable breadboard, which is any FPGA board ever manufactured, can be hooked up to our board. That means if you learned how to write hardware, hardware description language, you can write hardware description language, connected to our board in a meaningful way, and use that experience that you learned in college of engineering to add to these boards, in a way that a creative person would. And that leaves it open to everyone that wants to can be creative with our SXB’s.
KS: Nice. We’re talking about in embedded intelligence in the universe, and I thought we’re going to get into a theological discussion there for a minute.
BM: No, but we can get into philosophical discussion. By the way, I’ve been an advisor to Notre Dame for 20+ years, and I just recently looked at some of their curriculum, and I noticed from their junior and senior year, an engineering student can pick either philosophy or theology. So theology is the study of religion. If someone wants to study religion, I highly recommend the book by DK, it’s a publishing company out of the UK, that’s called “The Religions Book: Big Ideas, Simply Explained.” If you want to know about philosophy, the same publisher, DK publishing, has “The Philosophy Book: Big Ideas, Simply Explained.” If you want to know about economics, “The Economics Book: Big Ideas, Simply Explained.” If you want to learn about science, you can get “The Science Book: Big Ideas, Simply Explained.” So what I’m getting at here, the theology, I’m not a theologist. If you put me in a category, I’m a philosophical theorist, which means that if you see a bug flying around, or a bird flying around, or a bug crawling, or a tree that has leaves on it, in the summer, you gotta know that they figured out how to do that. And I’m calling that embedded intelligence. So if you say, well we already know about 50% of all the living things on planet Earth, and with all the scientists trying to figure out what the next interesting living thing is, we’ve only called covered half of them. So what I’m saying is there’s a whole lot of embedded intelligence. So if I were… some people think I was trying to start a new religion with this idea of a philosophy of embedded intelligence. I’m saying no, philosophy is a questioning. So when I say a philosophy of embedded intelligence, I just got a whole bunch of the questions that answers. I think we can find them. That’s what this country was built on. That’s why we have freedom of speech. That’s why we have freedom of religion. These guys that created our country, or founders of our country, like Ben Franklin, like Thomas Jefferson, George Washington and these guys. They called themselves Deists. A Diests’ believes you can know, and I add understand, God through reason and nature. Look up Deism. Just see it. And it’s the symbolism on the back of our dollar bill. So what I’m saying, is when you say “in God we trust,” the only God we can trust is all of us working together, and that’s my definition and you can find that online too. So what I’m saying is I’m not a theologist. I’m saying that Joseph Campbell wrote grope great books on comparative mythology, and I think his writings are really helpful in understanding that the theologist is trying to understand their own religion, and I’m saying that it’s easier to look at all the religions in that book, the religions book, just read that. You can see every religion that they could figure out when they published it, and is built by a whole bunch of people that are really uneducated, and they’re listed in the front of the book. Something in two or three pages you can figure out what Christianity is about; you can figure out what Judaism is about; you can figure out what Buddhism is about, which is actually a philosophy; you can figure out what Hinduism is. There’s some magic in religion. And I think they’re all good. I think there’s some magical mythologies and everybody that’s anybody from an indigenous further, all has their mythology. A mythology is an answer to questions’ unknown. Scientists try to figure out things out, and that’s where Galileo ran in with the trouble with the Catholic. The point is, that a scientist is looking for answers to the philosophers’ questions. Then, I’m trying to understand how to apply that. So as a result these boards are supposed to be a foundation for understanding of embedded intelligence in all of its forms. If you get online and look at the little… what you call it, slideshow that Dave Kramer helped me with four years ago, you can see I’m comparing natural to man-made. The top of the slide is natural. The bottom of the slide is man-made. My microprocessor was used in all the bottom. Like pace-makers and all these other things. So look at the complication of replacing a natural pacemaker cell, that every one of our hearts have to start with, and replace it with something we build because the actual cells stop working right, it’s very complicated. And St. Jude’s is one of the licensees that use my microprocessor, and they acquired three or four my licensees, one of them was Siemens pacemakers. So I’m saying I understand that my embedded intelligence technology, in the microprocessor form, can enable and empower other creativity, that hopefully is beneficial for mankind.
KS: Awesome. What haven’t I asked you that I should have?
BM: How old am I?
KS: How old are you?
BM: 70. No, just kidding about that.
KS: Are you going to keep going? Okay so you’re 70 years old. Are you going to keep doing what you’re doing, for the time you have left on the planet?
BM: Yeah, my goal is to live to be 100. I want to beat my aunt, who was my fathers’ sister; my father’s only sister. My father didn’t have any brothers either. So she lived to live 100 plus about two or three months. So I’m going to try to beat her. So that means I have about 30 more years to do things, that I love doing. But really I retired when I was 33. And what I call retirement, is doing what you want to do. And that’s what I’ve been doing. And if I didn’t do it, then you got to ask yourself, “why did you do what you wanted to do.” So that’s one definition of retirement. So I’m never planning on retiring from this technology. And I’m never planning to have this technology not available in some form. Now you got to know, that with our programmable logic boards, you have PMCU’s, I think that we’re the only ones that are supplying PMCU’s. However, others are, they just don’t call them that. A programmable microcomputer unit. Now my definition of microcomputer is a microprocessor, I/O and memory for data, and memory for program. So when you build a microcontroller like an Arduino, you’re outfitting a microcomputer with some special circuitry, to interface with maybe some analog-signals like sensors, and some digital-signals. So we don’t really think of ourselves as doing that. We think that the analog stuff should be put on it. So we will have shared projects with Mouser’s project -hare website, where we say okay here’s an A/D converter that you can hook up to the memory bus of our part, or hook it up to a 22, and exercise the data bus with one port, and the control boards with the other port, and that’s what people have been doing for… 35 years, maybe longer. So what I’m getting at is that when you look at it at the in a hardware description form of it, my foundation to make that technology available forever; that’s what my plan is. So I already have the PMCU’s. I already have the programmable boards.
They can… PMCU’s can be loaded into anybody’s… any fabricator of programmable logic devices, like FPGAs, so any board can use that PMCU, so you can say that for all eternity, the 6502 and 816 in programmable microcomputers, will always be available to mankind. Now with that in mind, you can go… if we ran the 816 as an example, on the most advanced processor that ARM is, or Intel is running, their processors on, we think we could run it 10 GHz, because we only need 8-bits of data. And everything is running on the chip so we don’t need any cache. So you can have the full memory space of the 816 in what would be considered all-cache. So what we’re saying then, is that if one was inclined to do it, we can run as fast or faster than any ARM processor or any Intel processor, and we will always be able to do that because the manufacturers, whether it’s Intel, or Global Foundries that acquired all of IBM’s semiconductor operations, silicon on insulator, would be an example of something that run faster than anything Intel has or that ARM is running on, we could do that with ours, because we have a simple processor that get the job done. So if you just take your calculator, and look at 10 GHz running this opcode, what could you do… what could you couldn’t you do… that’s the question, what couldn’t you do. And see one of the thing that’s important is to recognize, is that Microsoft, working with IBM and Intel, really define a platform for software engineers to work on. That’s why use Windows. It’s ten times as many systems out there than what Apple has, as far as Macintosh. Macintosh is a wonderful system. You think that I was a big fan of Apple, and I’m not. The reason why I’m not, is because they’ve killed off the Apple II, to make room for the Macintosh, and a lot of disappointed people followed that. And they could still be selling a product with the Apple II technology, and the 16-bit processor, they could be doing that better than ever, and they chose not to because they wanted to make more money, on the 68000, which became the PowerPC, became the x86 from Intel. Now there’s another example. If you need an example of why didn’t do the 32-bit just look at Motorola, IBM, and Apple, doing the PowerPC, and at the end of the day after all those billions of dollars invested, after all those thousands and thousands, or millions of hours, by designers, Apple chose Intel x86. So actually, the PowerPC was a prototyping platform for the Intel processor. That’s why I never did the 32-bit. So if you want to use the PowerPC, or a MIPS processor, or a Cparc processor, after what’s his name bought it, I’m saying you got some examples of what could have been, would never would have been, because of the marketing and the capital pressures of big business.
KS: I think this will be my last question. You mentioned using the 6500 hardware forever, so, people who listen to this podcast are still using their Atari computers was 6502’s, and if you could send them a message, and you can right now, and you what would you tell them?
BM: Well I’d tell them thanks for the joy of working with this technology, it was designed with a lot of love, a lot of care, a lot of heart, and I think it continues on with people that want to use it. And so I would say that unfortunately Atari, being acquired by Jack and I don’t know what Atari is doing anymore, Commodore that went away, was another thing that Jack got involved with, and the Commodore… I saw a Commodore smart phone recently, with Commodore, I don’t know what it’s called on it, just like an Apple logo on the back of the smart phone. So the brand name has some value to it, but the real problem is that most of those people don’t have the, I don’t know what you want to call it, the camaraderie thing, the love that’s needed to… I don’t know support the market interests, but the Atari, I think could come back as probably, whether you know it or not, 816 was designed because the ‘cO2 has an extra cycle and decimal correct which the asteroids broke up, so Atari never use the 8cO2, but they could have used the 816, because I found another way of doing decimal-correct with the same timings, so the timing of the 816 would allow the asteroids to not break up at the 1 MHz clock rate. So actually if you took the 816 and built a new Atari out of it, with something that would be interesting, and you look at the cost of that, see so most people want to compete with, let’s say the Arduino two-dollar value, which is a not-for-profit company giving away things, but they have to give them… they have to assign a purchase price to them, or the person getting it my just look at it and say “that’s nice” and put it in the in the drawer, and never use it. So what I’m getting at is that for those Atari folks, I would love to work with somebody that had the brand, and really wanted to come out with 16-bit Atari that would do some fun things. But it has to be done with the state-of-the-art technology. You could build a brand-new Atari with the real… if someone had the brand, you could build a brand-new Atari, with the fun legacy stuff that is what people like playing around with, or doing things with. And if you did that, I think you would have a following. Because there is a love for the technology. There’s as much for love for the Atari brand as there is for the 6502 brand, in my estimation. So, I cheer for Atari. I know Nolan Bushnell. He’s the one that gave me the award for… on the 25th anniversary of the microprocessor, I got the first award; Chuck Peddle got the second award, and then the other seven awards… the other five awards were given out I guess for the other five processors. But Nolan Bushnell is very creative, was very creative, and when in 1996 when that award was given to me, I talked to Nolan when he is interested in, what he was interested in, and what he is working on at the time, was multiplayer games. And I thought multiplayer games, what one game is enough at a time one player at a time? But that’s what we’re doing all over the world. We have multiplayer games. So Nolan Bushnell knew what he was doing in 1996. But see, I think it take somebody that has a vision, and a dream and bringing Atari back to life with the 16-bit 816 is what I would enjoy helping them do. But that comes from the love of the technology, not from somebody wanted to make the next billion dollars.
KS: I wonder if this is about what you’re talking about. This cartridge is called project Veronica. And this cartridge has a 65C816 S CPU, clocked at 5 MHz and 64K of RAM, and this plugs into the Atari 800 one of the Atari 8-bit machines and let you run 816 software on it.
BM: Yes, so then you have to right, Veronica… I’ve heard of that, and I forget… A woman designed it right?
KS: I don’t know. I was going to tell you what country it came from, and it’s somewhere in Europe I think, I’m not sure.
BM: See that, to me, again this is about Atari, you asked me the question… and I’m saying, I have an Alfa Romeo Spider that’s 26 years old. When I was a kid, I got a ‘29 Model-A, 1929 Model A, and it was about 1962, when I was in high school. I got that back on the road, and that’s wonderful. But I never thought of dropping in, a 409 engine into it. I didn’t think I was going to do that. And I didn’t want a kit car, which would look like a, whatever like a Ferrari, I didn’t want to kit car that had a Volkswagen engine in it. I like the real thing. And so when you add in a new engine into an old Atari system, that’s what I would consider a hobby market. What I’m talking about is a real market, and that is if you took the Veronica, and said “what would be an attractive case that would have the Atari brand on it that was actually sold by Atari?” And I’m saying that is possible, but you have to acquire the brand. Maybe you could work… somebody could work out and negotiate rights to use the Atari name, and do a real Atari game with the 816 in it, and you take that card… what’s the difference between that cart and the whole freaking Atari? I’m saying the only difference is the display interface, and a keyboard. And I’m saying, just do it. I don’t think having a cigarette case, or whatever you call it, a little… mints come in, or chewing gum…
KS: Oh, Altoid tin…
KS: Nice. Thank you Bill.
BM: You’re welcome. Thank you. Thank you for inviting me to talk with you today Kevin. By the way just in case you want to do it, if you want to have another one of these, I’m open to as many of these as you can stand.
KS: great. Thank you. What will happen… Here’s what will undoubtedly happen. I will post this in a month, and then I will get 10 or 20 emails from people going why did you ask him about this or that or the other thing. Then I’ll make a list, and then maybe we’ll have another conversation.
BM: That works for me Kevin. You hate to have an unanswered question don’t you?
KS: Thank you Bill, I appreciate your time.
BM: You’re welcome. Thank you Kevin.