Electronic Design

Gilbert/Hoff/Pease Panel Discussion Continued

Three design titans—Barrie Gilbert, Ted Hoff, and Bob Pease—give their take on where the EE profession is headed and why.

Here is the continuation of the panel discussion among Barrie Gilbert, Ted Hoff, and Bob Pease. The first part of this conversation appears in the Engineering Hall Of Fame section of our September 13, 2004 YOUR Issue, ED Online 8684. Electronic Design brought together these individuals-three of this industry's foremost designers and most respected thinkers-to participate in an informal discussion about the future of electronics engineering in America. We're sure you'll enjoy the rest of their informative, enlightening, entertaining, and sometimes profound conversation. Hoff had to leave early, but Gilbert and Pease talked on. See what they had to say.

BARRIE GILBERT: Not to dismiss these other questions, but let's get back to the technology. There’s another point of departure that I think is also important. Up until now, it has been assumed that the monolithic implementation of everything is the only sound way to go. If you submit a paper to the ISSCC bearing the title “A Monolithic Implementation Of A Very Hard Thing To Do,” it’ll get accepted. But if it says, “A Hybrid Implementation Of A Very Hard Thing To Do,” it won’t. And that’s dumb!

Hybrid implementation is being used again because we’re at a point where automated assembly techniques are now affordable and the technology is advancing rapidly. At Analog Devices, we make parts that contain a large number of individual elements on a tiny little header that looks like a regular IC and sells at IC prices. But inside the package, we can combine a bit of gallium arsenide with a bit of 0.13-micron CMOS along with a bit of complementary-bipolar signal-processing, and other components that are hard to integrate on a single chip. This is no longer an expensive technology. The days of military utilization have given hybrid assembly a bad name, but that’s going to change. We’re going to see an end to the long war between whether “CMOS is better than bipolar”, or “digital is better than analog”. They’re all part of the same world of technical challenge.

Clearly, if something can be done just as well or more robustly in digital, it darn well ought to be. The turf wars and the old partisan views of technology are at last being replaced by a broader and wiser point of view, in which we use the technologies where each is optimally suited. Then we put them all together at very low cost under robotic assembly, just as we would today build a monolithic circuit. Sometimes that doesn’t work well, for obvious reasons. If you want to access cache memory at high speed, you’d better not put it on a separate chip. But that’s a technical issue. It doesn’t mean everything has to be done that way. On the other hand, it doesn’t mean that everything has to be done monolithically either.

TED HOFF: You just touched on one of the main advantages of monolithic technology-minimizing the capacitance associated with interconnections from one chip to another. The issue we kept running into before was how many connections we were comfortable putting on a chip. Generally, each component that we're going to use has to have some number of connections. If we can come up with packaging techniques or interconnection techniques that allow us to do it without trying to print everything on a single piece of silicon or some other substrate, why then maybe we can do some more interesting things there.

Electronic Design: In terms of these technological trends, monolithic versus hybrid processes, how can the U.S. maintain or regain its technological edge? What are some of the scientific things that we need to stress here to be at the leading edge of those developments? And as part of that response, what are the sciences that need to be encouraged? Ideally, what would you like to see as the course of study young engineers should take to help solve those problems?

BOB PEASE: Young engineers have to get older and learn as they go. Young engineers can't always contribute, but the smart ones will pick up the ideas real fast, and they're valuable. As for the slower ones, well, we can always tolerate a few slow engineers. But thinking about the future, it doesn't all have to be scientific. It has to be thinking about what the customer really needs, as opposed to what the customer says it needs.

HOFF: One of the things that concern me is the lack of public policy that says it's even important. You get the impression that the only concern of our politicians is their next election. Telling people, "Hey, we have potential problems," doesn't get them elected. Everybody sweeps things under the rug.

PEASE: Charlie Sporck \[the National Semiconductor CEO who shaped the company in its early days\] had a good comment 10 years ago when he said: "We need to have an industrial policy. If we don't have an industrial policy, we're all going to end up shining Chinese people's shoes."

HOFF: We have issues beyond electronics too, for instance, our use of fossil fuels. We should have some kind of a policy that says our goal will be to build vehicles and all the rest that can run on something else, whether it be alcohol or hydrogen or whatever. I know people have been saying this for many years, but presumably and eventually we do run out of fossil fuels.

GILBERT: Yes, one of the things that always surprises me is that American elections are based entirely on personalities, and there is never a definitive statement of new initiatives—a manifesto of intent. In Britain, where I’m more familiar with the politics, the candidates are obliged to put on paper what they’re planning to do. Here, the politicians can wriggle out of any commitment because they never articulate it in full detail.

Electronic Design: And if you were helping to construct the manifesto for the engineering community, what would you like to see from the world of politics? What are some of the things that you would want to include on that?

GILBERT: Well, I'd like to see the same sort of support as the tobacco industry gets.

HOFF: You know, the last time I think there was actually a scientific program was probably when Kennedy made the announcement about putting somebody on the moon within a decade. And that kind of activity with a goal in mind would be kind of nice to see our government have-some kind of goals in terms of maintaining our technical superiority.

GILBERT: Certainly Kennedy was a great communicator and able to get people to wake up and listen to him. On the other hand, it’s important to remember that the space program was a reaction to Sputnik. I think the American people understood very well, back then, that they were in a war against a nation that was likely to do bad things to them, and if they didn’t get smart in space technology, they were in deep trouble. Today, we don’t have a threat like that. Of course, now we have terrorism. But whatever is done to defeat terrorists will inevitably be based on technology, especially information processing, although, government agencies don’t seem able to get their act together on how to organize a new information system.

But going back to what we should focus on. Certainly Bob and I would agree that one of the shameful aspects of modern electronics curricula is that they really don’t spend much time on analog. A few schools do, but most give the impression that analog is no longer important. Rather, they stress that “the future is really digital”, and “if you want a career in electronics, you better focus on that”. I would like to see that attitude moderate.

HOFF: Along those lines, I've had a concern for years about the academic world. They're always trying to catch up to industry-they want to get the same kind of processing equipment so they can move the same kind of silicon that's being done in the commercial world. It seems to me they'd be much better off if they were doing what I'd call the theoretical work, like researching speech recognition or pattern recognition. Rather than trying to build a faster computer, how about looking at the algorithms that you're going to run on the computer so that you know how to do a design? For that part of it, again, there seems to be some kind of coordination needed.

I'm going to have to drop out of this discussion now.

PEASE: C'mon Ted, tell us what you're really thinking.

HOFF: Now you can talk about me behind my back.

GILBERT: Okay, what do we talk about? Let’s get back to how you make electronics exciting for the youngster. It’s a tough call, but I don’t think it’s an impossible one. First of all, the industry (not the government) should do a great deal more than simply provide financial support for universities. People like Bob and myself are involved in teaching at universities and raising enthusiasm, but it’s not all that common. Many people in the industry seem to feel they’re too busy to mess around with extracurricular activities of that sort, and sometimes it IS tough to find the time. I’m on the IEEE Distinguished Lecturer Roster and often get invitations to speak at places that I really can’t make it to. Even if it’s only a day’s journey away, to go that far just for a one- or two-hour lecture is a big strain on my schedule.

Sometimes we use video-conferencing, but we could make better use of CD-ROMs. We could stand in front of a camera and do our thing, and enthuse away at the whiteboard. I think enthusiasm is the key ingredient. It’s not altogether what you talk about, it’s how your eyes gleam when you talk about it—that, I think, is what inspires the student. I’d rather spend an hour smiling and laughing and glowing with enthusiasm about some current mirror than telling them about the most advanced circuits I know about. And I often do that.

PEASE: Definitely, it's awfully good to have face-to-face contact and beat them up if they don't have enough questions for you.

GILBERT: Well then, maybe we need to institute a new mentoring program, one where students come to us. We open wide our doors, and offer a program where people like Pease and myself sit down with a student or two for an hour on a weekly basis. The student just has to sign up and appear.

PEASE: Possible. We've had good luck down at San Jose State where not only myself but also a few guys from Linear Technology have given an hour evening lecture before the seniors' electronics class. This went over quite well. We really showed them A, enthusiasm, and B, face-to-face feedback. "Why do you say that, Mr. Pease?" "Because that's why we know that blah blah blah." That was very challenging, but it was very fun, too.

GILBERT: And it keeps us on our toes.

PEASE: Oh, it damn well does! I just got back from India where I gave lectures in four or five places, and that was challenging, especially the temperature. It was only 120 degrees in the shade in Calcutta, and there wasn't any shade.

Electronic Design: Speaking of India, let's get back to outsourcing and offshoring. Right now, the basic feeling here is that while the manufacturing of electronics has certainly moved offshore in a big way, design is staying here, for the most part. Do you see that shifting?

GILBERT: Absolutely.

PEASE: Well, not absolutely, but gradually. Because if you want a good design and you get it made in the United States, it's a good design. If you want a good design made in some other place, is it good? I'm not sure, but I've seen some that weren't.

GILBERT: Where are all the digital cameras designed?

PEASE: Japan.

GILBERT: Why do we believe that we're the best designers?

PEASE: We're good at many industrial things. I was writing a note the other night: What did Bob Widlar \[the well-known designer of the first monolithic op amp\] think about the 555? He didn't think a blasted thing about the 555 \[a ubiquitous timer circuit from the early days of ICs\], because it was commercial or consumer stuff. He only cared about industrial stuff. So there are places where industrial-grade design is good. For consumer-grade stuff, yeah, camcorders are pretty good, but do they really work that well, that long? I don't know.

GILBERT: One other thing we need to say about analog is that it is durable. I think that’s a key distinction. Analog circuits go on ’til kingdom come. I have a product that was released in (I think) 1975, and it’s still among the top revenue generators for my company. That’s really quite surprising.

PEASE: No it's not. We do the same thing at my company.

GILBERT: Well, it is surprising to someone who doesn’t know it. You know it. We need to make it more apparent that one of the delights awaiting young analog designers is that they can expect to see their products selling for a long, long time. And they won’t be pressed into implementing the latest spin on, say, a processor that requires another little bit of logic here and there. Frankly, like Bob, I find digital design absolutely boring. There’s nothing to it, by which I mean, it’s not electronics. As I said earlier, if there were some other medium that did the same job, it would be used. Logical processors are only incidentally electronic. I call them “epitronic” because, from the Greek epi, they hover above their electronic underpinnings, but apart from that, they have no essential connection with it.

So I think that the problem we’re talking about is how can we keep electronics alive, because I really don’t believe that digital design is about electronics. Some may think that’s a funny point of view, but it’s something quite different. Today, it just happens to exploit electronic elements, but who knows what lies around the corner—whether it’s going to be based on DNA self-fabrication techniques, or something such as carbon nanotubes? Frankly, I don’t think either of those has got a hope. I think silicon is going to be around as an information-processing medium for at least another 50 years.

PEASE: For analog signal processing, too.

GILBERT: Of course, I naturally assume that. First, because silicon is so abundant: 28% of the earth’s surface is silicon. The sheer abundance of silicon and oxygen and aluminum—all the main ingredients of integrated circuits—and the huge installed equipment base, and the knowledge base, are going to guarantee silicon as the medium of choice. That’s really not the question. The question is: How we are going to use it? The emphasis, as we know, in technologies for processors has been to keep pushing down the gate length. Now we’re at 60 nanometers, 90 nanometers, wherever it’s manufacturable—it’s somewhere down there. But this obsessive and near-paranoid emphasis on gate length and gate-oxide thickness is taking no account of the needs of the analog world, for which it’s totally the wrong stuff. We need to resurrect the notion that electronics is a scientific pursuit—a branch of applied physics, and all about the nature of the real, physical world, rather than the prevalent jaded view that “analog circuits are also one of the things that you can do with silicon, by the way”.

PEASE: I wish it were easy to propose a university for analog as a scientific pursuit, but actually if we keep growing our analog business as adjuncts of the processor's needs, we're still doing good business. Can we wag that dog? I don't think we can wag that dog, exactly. I think customers can wag the dog by saying, "I don't want one that's faster. I want one in which the battery lasts seven days." I think that'd be neat as hell. Even Intel has caught onto this concept that faster, faster, faster is not necessarily better, better, better. It may be slower, slower, slower, and yet it is a lot better.

GILBERT: By the way, this emphasis on speed and magnitude gets to an amusing point. Apple had its little iPod, which I believe stores 10,000 tunes. Sony comes along and makes the big claim that it can store 15,000 tunes. Now I don't think I even know 10,000 tunes, so what's the point?

Electronic Design: Instead, battery life should be the point. What you really want to know is "How long can I listen to these tunes?"

PEASE: People have been neglecting battery life because the faster, faster, faster you go, of course, then you need a better, better battery. Pretty soon you run out of juice and you can't run the thing for an hour-and-a-half. And, no, you can't do video games at the full 87-Gigabyte/µs speed. You have to basically give up your battery life. People have learned that's not such a great tradeoff, is it?

GILBERT: Going back to how should the profession change to survive, I’m not sure it’s entirely in the hands of America. In fact, I’m darn sure it’s not. The world is a big place. We’re only a tiny bit of it. What are we: 3% of the world’s population? There’s a big world out there, with a lot of people who are very smart, who are increasingly aware of the benefits of modern life, and who are grasping it with both hands. That is going to be the driving force behind the economic situation in the coming decades.

Electronic Design: But the United States has been driving the advance of electronics technology for decades and decades now.

GILBERT: Well, maybe that’s going to change. I mean, why should it go on forever?

PEASE: Barrie, let me ask you a question. What kind of computer is the best-selling computer in India or China? They don't need 87 giga-anythings per second. They need a pretty good cost-effective one. I'd be curious to know. I haven't looked into that.

GILBERT: I would guess they buy what's available, and what's available are high-speed computers these days.

PEASE: Yes, but they might buy last-year's high-speed computers, which are still in production. If that's the case, they're getting a cost-effective thing. Man, for a high-performance video camera, you can look at last year's model, and it looks pretty good and the price is cut in half.

GILBERT: By the way, it’s good to bear in mind that most of these computers are no longer made here in the ’States, anyway.

PEASE: Well, that's largely true.

GILBERT: I think we need to come to face to face with the reality of the situation, which is that America is a nation headed toward decline. I don’t see any other way to see it. I believe we’re in a moral decline too, but that’s another story. I sense we’re in a decline in terms of true-grit leadership, at many levels, and I don’t think there’s any measure that any politician may take that can avert the inexorable slide into insignificance. Sorry to say that.

PEASE: Well, I think Mr. Kerry is going to try to do that and he still has several months to make his positions clear, which of course he has not yet made clear. He could slow that down if he's really damn smart, and maybe he can. I don't know.

GILBERT: I don't think so.

PEASE: He can try. He has to try.

GILBERT: Sure, but it's one man against the tide.

Electronic Design: Well, from this conversation, we're saying that it would take a coalition of government, industry, and universities.

GILBERT: Right, a global coalition.

Electronic Design: Another important point you are making is that we have to understand how to interface with the rest of the global community and maybe redefine our position within that community to continue to grow here. It may be a matter of connecting and understanding and making those relationships happen by reaching out to the rest of the world to share our expertise. Do you see the U.S. as continuing to be the leader in analog technology around the world?

GILBERT: Let me answer that in today’s terms. My company supplies the major fraction of the analog components and data converters for digital cameras, and a lot of other display devices as well, so right now those parts are indeed coming from the U.S.A. Japan seems to be peculiarly shy of analog design, and I’m not sure why that is. Maybe it’s a reflection of the trajectory that young people have taken there, with an excessive emphasis on digital. Who can say whether we will continue to be top dog in analog? A lot of our IC designs are being done in India right now, and their designers are doing very good stuff. They are in that fortuitous position of not knowing what’s impossible, so like the bumblebee, they manage to fly even though everyone says it can’t be done. That’s a distinctive feature, which needs to be noted: It’s sometimes good to be a little ignorant.

Again, China comes to mind. It surely won’t be long before China has a democracy of some sort. I can’t see them staving it off. You can’t stem the tide of progress in that regard. What kind of a democracy it’s going to be, we can’t foresee. Whatever it is, I think the Chinese people will be enabled or empowered by it to do even more in the world scene than they’re doing today. It’s a very large population, and once that population gets the hang of technology, as it is beginning to do, the world is going to change in a very major way.

PEASE: People begin to get the hang of analog circuits, Barrie? I don't know. You only need 20 people to get the hang of analog and there we go.

GILBERT: Well, there are probably 20 people in China.

PEASE: There are probably 20 really brilliant people in China. If we ever let them have their way, they'll be damned smart analog circuit designers.

GILBERT: And hard working. I think that’s another issue. America has become fat and complacent. Maybe that’s at the root of our problems. We’ve been so used to a cornucopia of plenty that we can’t get used to the idea of using less: less gas, less energy…

PEASE: …and working hard.

GILBERT: And working much harder than we do.

Electronic Design: What would each of you say to the young people today to get them interested in electronics engineering as a career? How would you motivate them?

GILBERT: Principally it would be to illuminate the kaleidoscopic variety of electronics and not leave them with the impression that electronics is all about video games and communications. There’s more to it than that. Somehow I would like them to see we’re in a process that started about 100 to 150 years ago. We’re part of a continuum, and they can contribute to making the future. I like to say, “There is no future—the future is what we make it." In fact, there’s no present, either, because the present is that infinitesimally-thin membrane between the past and the future, and we’re within that membrane all the time.

Anyway, the future is something we make, so I try to get students to understand that they have the power to change the world in some way. People like Bob and myself can see things we’ve done that, maybe only in a small way, made the world a bit different. I don’t think young people are going to have much chance at doing that by being one of 100 people chopping away on the next billion-transistor microprocessor. Analog still offers the opportunity of being a one-man band. I think Bob would agree that most of our older analog engineers are maverick individuals with a stubborn and singular vision for “what might be”. And if our new designers have track records that are halfway decent, they’re also allowed to pursue their unique vision, and make things happen that no one thought could happen. So I would try to emphasize those themes, and not really worry too much about technicalities.

PEASE: Okay, Barrie. What's 12 times 24?

GILBERT \[answering instantly\]: It's about 288.

PEASE: That's right! You have no idea how long it takes to pry that answer out of 1% of the audience in Tokyo. Part of the problem is they don't sell eggs by the dozen in Tokyo. They sell them by the 10-packs, which is perfectly logical except you don't learn your 12-times tables in Japan. It's not easy these days for kids to learn math because they're all taught to just learn the minimum amount and then use their calculators. They don't have to learn any math tables. But if kids are good at math, they may be good at science. Encourage them, they may be good at engineering, they may be good at electronics, they may be good at analog, they may be good at digital stuff. It's a big fan-out of all the things you could possibly be interested in. We have to encourage the kids who have good math or scientific or mechanical backgrounds to get into things where they can have fun and get entranced by the science.

GILBERT: For people like us, and for the purpose of this conversation, what we’re striving to do is to keep the analog flame alive, not just because we like the color and the warmth of it, but because we understand how important that flame is. One aspect of this multi-faceted problem is that analog isn’t sexy. I mean, carbon nanotubes and DNA assembly techniques and all that stuff are being promoted as the next new wave, which is manifestly rubbish. If we can turn up the analog flame and say, “Look, some of these new things may well work out, and it’s certainly very exciting to be at the forefront. But there’s still a forefront of electronics and it’s not in digital computers. That’s the forefront of information processing.”

The forefront of electronics will continue to be in analog circuit design. For example, radios are forever. We’ve only had radios for 100 years, for heaven’s sake. We’ve only just started. There’s a huge amount to be done yet in wireless communications. Think of all the things we can do with wireless sensors. I just bought a new house and it has a little kitchen weigh scale that’s built flush with the granite top and it communicates to a computer by a wireless link and gives you the measure of your flour or whatever into the computer by wireless link. I think wireless sensors of every imaginable sort are going to have a huge future.

PEASE: I could just see my wife trying to scrape and dust the flour off the counter.

GILBERT: The point is that there’s a lot still to be done, and somehow we need to get that message across. I believe there is a great cloud over electronics that says “old analog is over”. In fact, whenever you read of the latest advance, isn’t it always digital? I mean everything seems to be digital. What’s the best way of making music? CDs. What’s the best way of seeing videos? Digital Video Disc. What’s the best cellular phone? Digital. Sure, digital plays an essential role. If it weren’t for digital technologies, you couldn’t organize a cellular network for phones. But the heart and soul of a cell phone is the transceiver that makes it possible to talk at such long distances. Somehow we need to just make it more obvious that the future of electronics is as much about analog as digital.

PEASE: Well, I tend to agree on that one, because there sure is a lot of work to be done that's largely analog in results and largely analog in terms of internal practice.

Electronic Design: Somehow we're not surprised that you agreed on that one.

GILBERT: We don't agree on a lot, though, I have to tell you. Maybe we can get into a little fight about simulation.

PEASE: We'll discuss that later.

Electronic Design: Any other final thoughts you want to share before we wrap up?

PEASE: I'd like to follow up on my earlier point that supply and demand is a cruel mistress. Is there demand for digital? Well, it seems to be. Is there some supply? Yeah. What do you pay for? Well, you pay for what you get, and you pay for a surcharge for what you think you want even though you don't really need that junk.

GILBERT: There is one other thing that’s worth mentioning. The recent marriage between the radio and the computer has led to a new view of system design. And that’s a whole arena in itself. We’ve been talking as if we’re keen to teach analog cell design, which I certainly am and I’m sure Bob is, but there’s an increasing demand for systems engineers—people who understand how to marry the computer to the op amp or to the radio receiver. And it’s the integrated-circuit potential—the capabilities of integrated circuits—and their limitations that are shaping many new techniques.

For example, Bluetooth is a very easy thing to implement because it takes into account the technologies on which it is realized. In contrast, GSM is a very tough thing to implement because the people who sat around the table and talked about their Groupe Systeme Mobile hoped that it could all be done and they never worried about it. In fact, there’s a very thick book on the GSM system, and it doesn’t even talk about electronics. So this fusion, this marriage, this concord between the system needs and the way in which those systems will be implemented is stronger than it used to be.

PEASE: Good point.

Electronic Design: Any other points to make as we finish?

PEASE: Wait a minute! We have to say something bad about Ted, if we can, but I can't, so....

Electronic Design: Well, all this badmouthing of digital has taken advantage of Ted's early departure.

GILBERT: Quite honestly, I know that was a joke, but the discussion wasn't really badmouthing—it was a sincere expression of the problem. The problem being that electronics is viewed as digital and someone comes to the rescue with high-speed processors because you need them for speech processing and so on. The common view is that it's got to be done digitally and unfortunately that's not true.

PEASE: We agree. Hey, what a good place to stop.

Electronic Design: Gentlemen, thank you for your contributions today. We really appreciate your taking the time to participate in this really fascinating conversation.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.