Ray Holt’s career spans the Cold War and now, as a teacher in Mississippi, Ray is actively passing on his own education as an accidental engineer.
In this episode we hear from Ray himself about his early involvement with hardware engineering of semiconductors, as they were used by the military in fighter jets–and about their commercial adoption by tech titans like Intel and AMD in the present day.
Full text transcript below the fold:
- Ray Holt (Wikipedia)
- Ray Holt and the history of MP944/Cadc @ Rome Technopole, 2017 (YouTube)
- The Accidental Engineer by Ray Holt
- First Microprocessor (dot) com
- Mississippi Robotics (dot) org
Full Text Transcript
Max: Welcome, all. Max of the Accidental Engineer here. Today we are joined by Ray Holt. Thanks for joining us, Ray!
Ray: Hello, Max. Thanks for having me.
Max: For our audience that doesn’t know about your background you are an awesome guest to have on the show, and I want to thank you for coming on.
For people that don’t know you, do you mind introducing yourself?
Ray: The semiconductor was invented out of Fairchild Semiconductor in Palo Alto and then Texas Instruments invented the transistor. And then when I came along I used that technology to make a microprocessor out of the technology.
We had a set of chips that actually worked as a computer then.
Max: Do you mind sharing for our audience why you were working on this stuff at the time?
Ray: How about if I lead into it with how I got started? Because that’s an interesting story.
Ray: Out of high school where a lot of us get guidance the only guidance I was given was, “Don’t go into engineering.” And that was because my mechanical aptitude was very low.
Mechanical engineering was a real hot field at the time, so I was advised not to go into engineering. But even though my math was high, they didn’t really key on that.
I just kind of floundered for at least a year at a community college. I actually walked out of my classes the first semester. Got all F’s. Came back wanting to make the baseball team and couldn’t make it because I had all the F’s.
I ended up transferring to the University of Idaho where a friend of mine was quarterback. Surprisingly, they accepted me. I had to declare a major. I knew it wasn’t engineering, so I looked on their list. I saw Forestry. I liked the outdoors, so for three years I studied Forestry.
The dean of forestry called me in one day and said if I wanted to stay in forestry I had to really do better in chemistry, which I had never had in high school. But he suggested I take a class in physics or electricity. And I did and it was heavily math, heavily hands-on, and I absolutely loved it. I got straight A’s in everything. Never dreamed I’d ever get all A’s in a class. Homework, finals, midterms.
And then I realized I was surrounded by engineers. I said, “This is crazy. I think I can be an engineer.” And so I started looking around at engineering schools and found Cal Poly in Pomona, which had a new program they were starting called electronic engineering. I think it was 1-year old at the time. It was going to concentrate on small circuits and not the big power circuits, as electrical engineering does.
They accepted me. I transferred. I had to do another three years of college and absolutely loved it. Maintained a high B. And near the end I had to take an elective.
Looking down the list I chose an elective that talked about math and computers. It was called Theory of Switching Systems. And I loved it. It was basically binary arithmetic and logic design. But, they didn’t really apply it to the design in computers. Just designing control systems in my site.
When I graduated I had three [job] interviews. The two I liked didn’t make me an offer, and the one I liked made an offer, which was pretty good. $17,000 at the time. That was fairly high for a bachelor degree engineer.
And they wanted me to design amplifiers, which was one of my worst courses. But, it was an offer. And I accepted it.
The first day I walked in the Human Resource gentleman said, “Well, I see you’ve had a computer class.” Which stunned me, because I didn’t know I had one. He goes, “Yes. Right here. The Switching Systems.” I said, “Oh, yes. Yes.” He said, “Well, you’re the only person in our engineering department that has had a formal computer class. We have a new project for you.”
He took me to the basement of the company, which is where the production lines are located, and he took me up to a big box and took the lid off. He said, “Do you have any idea what this is?” And it looked like a transmission, but it had a lot of wires around it. So, I just said, “It looks like a transmission, but I don’t think it is.” He goes, “You’re right. It’s a mechanical computer.”
Full of gears and tams and potentiometers, wires. He said, “This computer is used on the F-4 phantom jet.” Which was our main jet flying to Vietnam. He says, “Your new project will be to convert that computer to solid state or electronic computer.” That’s what they called it at the time.
Well, since I was told not to go into engineering because of mechanical aptitude, I wasn’t quite sure this was gonna work out.
But, of course, I didn’t say anything. I was very excited just to be there.
I was the youngest in the department surrounded by amazing, amazing engineers that really knew their stuff. Very specialized. And I ended up being one of two specialized logic designer on the project.
The good thing is the next day they said, “Well, we really don’t have the contract yet. We think it will be two months, so just start studying. And if you need to go to a course, just let us know. You need to buy books, let us know.” And that’s what I did. And after two months I was pretty savvy on designing computers logic, you might say.
Max: For our audience that isn’t familiar with where Cal Poly Pomona is located, it’s in Southern California, nearby Los Angeles.
What was the job market like at the time that you were graduating Cal Poly Pomona and finding these jobs? Were all of them located in Los Angeles?
Ray: There was a huge amount in Los Angeles, because of the space program. And most of them had some relationship to spacecraft. And the company I hired into, Garrett AiResearch, they did a lot of the propulsion systems, engines, and air pressure systems on commercial and military spacecraft.
The three I interviewed with actually came to the school. And, at the time, I was happy to have three looking at me, but it was a good market for engineers. You know, I think I probably could’ve gone anywhere in the country if I would’ve pursued it. But my family was in LA, you know, so everybody likes to stay near their family when they first start.
Max: Certainly. Jet Propulsion Laboratory or JPL still is in Pasadena, I think.
Ray: Still in Pasadena, yeah. Later on, I was able to do a computer interface board for them for the moon lander. And I got to go there and actually see the prototype of the moon lander. It was pretty amazing.
Max: On your first or second day on the job, that’s a pretty overwhelming experience or an overwhelming amount of responsibility that must’ve felt like they were giving you.
When or how did you know that ultimately the work you were doing would end up on fighter jets in the real world?
Ray: That’s a good question. It took about six months, because they didn’t really talk about the real purpose of the plane. They just gave me, like, a specification. Contract specification. So, the computer had to do this, this, and this. Now, the amazing part was, I found out early on I didn’t have to know anything about mechanical. And we had a gentleman, Bill McCormick, was his name. He was an applied mathematician. And I remember talking to him about all this mechanical stuff. He says, “Ray, don’t worry about it. The mathematics used in the mechanical computer is the same math used in the new one.” He says, “We’re gonna go by the math.”
So, all I had to do then is to start thinking about how to implement math equations, which is all a computer does anyway. You know? Plus, the controlling of the input and output. And then sometime around six months, they said it was called the F14 and it was going to be a supersonic fighter jet. We all know there was a Cold war, and you know they told us ILO is supposed to be the big jet to counteract the Russians. So, that’s kind of how we proceeded with it. The project has a lot of attention in the company and with the Navy. And a lot of demanding that we stay on schedule. Every time I would ask my boss if we could have a week delay or two weeks delay, he said, “Absolutely not.” He says, “Do you need more people? Or do you need more money?”
And that’s kind of how the project proceeded. Ultimately, we finished on schedule. Almost exactly two years from when I was hired. And the computer worked the first time, and we got to deliver it to the Navy. And the plane flew ahead of schedule, and was very successful. Now, there’s a lot of interesting engineering in there. I’ll just kind of highlight it.
Ray: Making a computational computer on small chips had never really been done before. So, we worked with one company called American Microsystems in Santa Clara, California. And they helped us come up with a design specifically. Now, it took three iterations. So, we would design what we thought was gonna assist them that would work. They said, “No, the chips are too big,” or “We can’t go that fast.” So, we’d go back and design again. And they said, “No. Why don’t you think about doing it this way.” So, finally the third time they said, “Yeah. We think we can make the chips.” And that’s kind of where we got to the point where we knew we had a working design. We just hadn’t seen the chips yet. If the chips were going to be larger than most that have ever been made, more transistors on them, because it was a Navy contract you had to go out for multiple quotes. And so, we sent quotations out to Rockwell, General Instruments, and American Microsystems.
All three of them came back and said, “We don’t want to make them.” Including American Microsystems. They said, “Yeah. We helped you with the development, but you only wanted 1,000 total and we’re making 10,000 calculated memory chips a month, so we’re not interested in doing the production.” And so, that was a shocker to all of us. And so, we told the Navy that. Their administrative people there. I never really directly interfaced with the Navy. And the project was halted for two weeks. And all of a sudden someone came to me and said, “If you could pick one of the three, which one would you pick?” I said, “American Microsystems, because they already worked with me in designing the chips.”
About three days later, they told me, “Okay. We’re going to continue the project. We want you to fly out there the next day from Los Angeles and start working on chips.” And I had no idea what happened. I found out later that the president of Garret AiResearch, the company I was working for, flew out there and made an offer to buy out American Microsystems, 51% of their stock at a very good price. And so, my company owned them.
Part of the deal was they would help us set up a second company, The Second Source, because all military product had to have two sources. And so they set up a new semiconductor company in San Diego called Garret AiResearch, which later on was bought out by Motorola. And then after the chips were made the stock deal was that Garret would sell the stock back at a very low price. So, the three owners of American Microsystems probably around 35 years of age ended up becoming multimillionaires just from this deal.
So we proceeded and we were able to work with them and lay out the chips and we did exhaustive, exhaustive testing of the binary and the math. And first set of chips came out working. It was absolutely amazing.
Max: The history of the time exclusively focused on aerospace and aeronautics. Were any of your coworkers or yourself suspecting that what work you were doing now would have commercial applications outside of the Vietnam War and the Cold War?
Ray: The answer is yes. My entire company was aerospace related. And there was a lot of vendors. Most of LA, with Northrup and North American, LA was full of aerospace and San Diego. And, you know, a lot of the technology from the space program filtered down to the commercial aircraft. And then eventually into our normal day to day. A lot of the jobs were related to aerospace. And because of aerospace, that really propelled the semiconductor market. The making of small chips. The making of small chips that did calculating. You know, it changed our calculator market from mechanical calculators to electronic. And then eventually the computer market.
Now, what I described to you all happened from 1968-1970. June of ‘68 to June of ‘70. And computers were only in large companies. You know, the government had a large computer and every insurance company had one. And that was about it. The thought of having a computer in your home was just outrageous. Why would you want one? You know?
That wasn’t in the thinking, but having a piece of equipment that did calculations fast was very desirable because as our needs grew we wanted to control big equipment. Control nuclear plants or control manufacturing plants, and you needed controllers that worked fast and have a programmable microprocessors was really a big advantage. That didn’t kick in until the late ’70s. So, when I finished this project in ‘71, they wouldn’t let me publish a paper on it. They wouldn’t let me patent anything, and they just said, “Just don’t talk about it.”
I never had to sign anything, but I did have a clearance. With clearances you just, being a patriot, you know, you just didn’t want to go out and blab when they said you’re not supposed to do something.
Max: Fair enough.
Ray: So, I never really revealed it, except to my family and maybe a good friend. But, it didn’t really mean anything to them either. So, life went on. And after the project was over. I moved up to Santa Clara and actually worked for American Microsystems for almost three years. And designed a couple of calculator chips and two more commercial microprocessors. And that was about the same time Intel started coming out with theirs. And Intel was a very big marketing powerhouse, so at some point, when the microprocessors were accepted in the industry, the Intel marketing just dominated everybody. So, my company in 1974 had a meeting between the owners and the marketing people and none of us engineers, and they decided there was no future in microprocessors. And they fired 25 of us.
Because they wanted to stay in the calculator business. Now, I can see from a marketing business point of view that might’ve been a good decision, but certainly, it doesn’t show that you’re a visionary. You know, you’re just looking out for yourself day to day. And microprocessors, the volume wasn’t that big. I mean, if a company bought two or 300, it’d be fortunate. Today, Intel claims they sell a billion a year. You know, they’re all embedded into equipment. But what was interesting is one of those 25 ended up going to starting the Atari computer line. One went to AMD and became the chief microprocessor designer at AMD. One went to Intel and became the head of the graphics division. One went to Commodore and ran their engineering. So, all of that knowledge just spread out across the industry.
And they had already gone through this high tech chip processing development. So, even though a lot of people tell me, “Oh, well your design was secret, so it had no impact,” well, it was probably true that the design itself didn’t, but what all the people learned from it had a huge impact on the industry.
Max: You, yourself, had a role in spreading some of that information and education. You were hired by Intel after this point in time to help educate the market about why this was a big deal. Do you mind sharing with our audience and little bit about what you did when you were hired by Intel and the types of people that you came into contact with that might’ve contributed to the massive growth and adoption in this technology?
Ray: Yes. In 1974, the Intel marketing group asked myself and my business partner, Manny Limas, to come over and join them in a meeting. And the meeting was basically what can we do to make this product viable? Because the owner of Intel told them they only had a certain amount of time or else he was going to drop the line. So, it was about a day discussion and I remember we … there were some options, but the big option was we needed to get out there and train the users. We wanted to sell to engineers, but engineers don’t know how to program. It’s a new concept.
The only programmers around were data processing programmers for insurance companies or government. Programming your control system, reading bits, and sending bits out that was a new concept. So, we came up with a one week program. A training program where we would teach two days on their four bit computer, the 4004. Two days on the eight bit computer, the 8080. And then one day the gentleman named Gary Kildall from Digital Research, he would come in and teach the high level language that he wrote for the computers.
We probably trained 800-1,000 engineers over two years. Keep in mind, we didn’t have, like, computer terminals, we had teletypes. We had to drag around the country. These are, like, 150 pound units. Big boxes. It was a real pain. And some hotels, like in Chicago, they refused to carry them inside. So, they would drop them off at the shipping dock and they made us take them into the conference room.
We ended up training those engineers, and the highlight was the one engineer … Well, two highlights. One is near the beginning the professor of my Switching Theory walked in and was going to be in the class. So, I had the privilege of teaching him how to program microprocessors. He’s the one that taught me logic.
Max: That’s incredible.
Ray: That is incredible. We had a lot of fun. But then the next year this vice president of NCR had about a thousand engineers working for him, was in the class and I think it was a Thursday, so about the fourth day he finally got his 20 or 30 line program working that detected the throwing of a switch and turning a light on. And he just jumped up, threw his hands in the air, and said, “Hallelujah!” And he realized that a design can be programmed. And if there’s a mistake you can reprogram it. You don’t have to redesign the whole board, which is a huge time saver. And he ended up going back and converting the whole company to microprocessors and was one of Intel’s largest customer for years and years and years.
Now, every microprocessor took memory. And Intel sold both, so that was really a great thing. And that’s what really helped Intel out a lot. You might sell one microprocessor, but you could sell 12 memory chips with it.
Max: Makes sense.
Ray: Yeah. I don’t know why American Microsystems didn’t see that, but they didn’t.
Max: One of the topic that’s come up previously on our podcast is the topic of credentials and getting licensed. Did you have to get any licenses to do the work that you’re being hired to do? Or was it a pretty open field as far as being able to obtain jobs without licenses?
Ray: Absolutely open. Licenses didn’t come until mid 80s. And that’s when Microsoft started pushing companies to have certified users and developers on their products. I even did some beta testing for some Microsoft license. And it was interesting, in the late 80s, I think, I tried to hire in or I interviewed for a company and they asked me what licenses I had, and well, I don’t have any, but I was the tester on a lot of them and I’ve been around 20 years. “Oh, I’m sorry we can’t hire you. You don’t have a license.”
Max: That’s pretty dumbfounded that they would be so short sided.
Ray: Well, it kind of showed that the market was to the point where they didn’t want to take a chance on your knowledge if it wasn’t certified. Because that kind of gets them off the hook, you know? Like, if I created a problem, they could say, “Well, he was certified and we did what we were supposed to do. It was all his problem then.” And that’s how it is today. Especially when networking came along. Companies wouldn’t hire you to mess with their network unless you were certified because you could really mess them up. Just by changing permissions wrong or changing passwords wrong. So, I understand the certification. But in the 70s and early 80s, nah, if you could design something that worked you could sell it.
Max: Did the ending of the Cold war and the dissolution of the USSR have a transparent effect, from your perspective, on the engineering industry of that time? This is in the early 90s.
Ray: I don’t think so. I know it changed the military a lot. By then we were so consumer oriented with handheld calculators, small computers, and all the electronic devices, there was plenty of engineering to go around. And plenty of programming. Programming was in demand like it is today. Now, what did go down was mechanical engineering. I mean if you graduated 80s, 90s in mechanical engineering, you probably would have a tough time getting a job or aerospace engineering. But if you were electrical, electronic, computer science, you’d have no problem.
Max: Now, today, in 2018 … It feels weird saying that, you, yourself, are a teacher and an educator. What do you tell your students about … Maybe high school students and college students you speak to about guidance you’d give them about entering the job market and pursuing an engineering career?
Ray: I’m no different than their parents and the other teachers. They don’t want to hear it. You know? But you have to tell them and I’m in small town Mississippi. I concentrate in rural areas, so there’s not a lot of vision. But you have to kind of create the vision on what’s out there and then you have to convince them that yes, if they take the right courses and learn the material that they could be out there doing that also. This is my seventh year here and I think next year I may have my first junior high or graduate from college. But there’s a lot more coming in in back of her.
Mississippi’s a little unique because the isolated rural areas, you know, the vision is really lacking. So, we’re having to work on that a lot. Do field trips and visit the colleges and companies. Yeah, the kids aren’t too impressed with hardly anything you tell them. But we know on this side, as adults, that if they don’t learn certain things it’s a pretty cruel world out there. And basic math, basic science, you know, consumer kinds of electronics, they really need to learn that. We’re going to live in an automated high tech world real soon. And if you don’t have the tech knowledge of that stuff, it will be hard to exist.
Max: One of the types of things that you do as an educator that, I think, is a really good activity for educating kids about opportunities in tech and engineering is competitions. Do you mind sharing with our audience some of the competitions that you have your students participate in?
Ray: Sure. Well, to facilitate the competitions I started a nonprofit called Stem Advancement. And in Mississippi we operate as Mississippi Robotics. And so, because I work with a lot of rural areas … Well, I teach two days a week at one particular rural area called Woodvale, Mississippi. It’s a Wilkinson County Christian Academy way down by Louisiana. But I travel to six or seven other ones every two weeks. I realize that these kids don’t really leave their community very often. I just said, “Wouldn’t it be great to get them together once or twice a year and let them compete with each other?” And they just ended up loving it. They love the challenge. They love meeting new friends. A lot of them get each other’s phone number and they text during the year.
So, that has grown quite a bit. In four years it’s grown from 35 per competition to about 250. And we picked up a major corporation, Nissan Automotive. And they sponsor the competitions, and they let us do one of them inside their factory once a year. And that’s coming up this April. And they realize that working with young kids in their robotics, math, science, automation area builds future employees for them.
And what’s really been nice, for me, is that in my discussions with Nissan, they’ve introduced me to, what they call, employable skills. Now, first, I thought that was, “Oh, math, science, history.” You know, being able to be interviewed. No, it’s turning screw drivers and realizing gear ratios. And they have a list of about 2,000 skills that are needed to run a factory. And they suggested to me, “Is there a way you can start teaching your kids these skills?” And it has nothing to do with the curriculum. But it has a lot to do with the activities within any curriculum.
And so I started a conference, a teacher conference. We call it the Stem Teachers Industry Trainers conference. And we have one coming up in two weeks. Our third one. And so, I get all these people together and we talk about these skills and how we can use them in teaching all of our various activities. And Nissan just loves that. Because in a normal math class or a normal science class, these skills may not be used. And now that we’re introducing them and teachers are using them, and we use them in our competition, they can see that we’re building employees and they really like that. So, just to have that relationship between industry and education is really unique, I think.
Max: Yeah, I think you’re doing a wonder of a public service to your community as an educator and trying to innovate with alternative, outside of the classroom, types of forms of education. So, I just want to personally tell you that I admire the type of work that you’re doing.
Ray: Thank you. Appreciate that.
Max: Before we sign off this episode, I want to mention that we’re going to include notes for the show with links to some of the stuff that Ray’s been mentioning. And also, I want to plug real quick that Ray is also an author. In which he’s written about his life experiences as an engineer working on the things we’ve mentioned in the episode. Ray, do you mind mentioning some of the works that you’ve authored?
Ray: I’ve actually only authored one book called The Accidental Engineer, which is how we got together. I love what you’re doing.
Max: We’ll include both of the links on the webpage where we post this interview. And, of course, to your book.
Ray: Great. I appreciate it.
Max: Thank you for coming on, Ray.