RW: Brian Halla served in a number of positions at Control Data and Intel before joining LSI Logic as Executive Vice President of the product group. In May, 1996, he joined National Semiconductor as President and CEO. Up until that time, he'd been involved almost exclusively with digital products. In this 2006 interview, Brian describes the many positives associated with the less glamorous but more profitable analog semiconductors.
RW: Tell us about growing up, your family, siblings - tell us about that.
BH: Yeah, I grew up in Fort Dodge, Iowa with three brothers and a sister and my dad was the resident FBI agent for thirteen some counties in the Midwest. And went to school at the University of Nebraska . Squeezed in a four year EE course into five. And started my first job at Control Data Corporation. I've had to have had the most fabulous career anybody could ever want. Control Data was time spent helping get the CDC 6600's into NASA so they could put a man on the moon. And we all watched that first footstep and -
RW: And that used discreet transistors. They're not -
BH: The transistors actually - those first mainframes used ferrite core magnets for the memories. And then these three young guys from Fairchild came along and integrated a bunch of transistors to create the first thousand one thousand bit DRAM. But my second gig is going to Intel in 1975 from Control Data which, of course, no longer exists, which was really kind of a government house, to Intel was the difference between night and day. It was everybody constantly bouncing off walls. The company was just a little tiny company then. And there's so much electricity all the time - all the time. And my first week on the Bob Noyce and Grove and Moore invited me to have lunch with them in a conference room. And I thought this is just unbelievable - unbelievable. But, you know, Intel was an experience I'll never forget. I spent fourteen years there. But it was always at general quarters. And that was just the way Andy ran the company. And people used to say if times were supposed to be good, how come it feels so bad? But obviously it got the desired results. And then 1988, moved over to LSI Logic to get the company in the microprocessor business. Met Rob Walker on the first day I think. And, you know, there was kind of, you know, if CDC was the man on the moon and Intel was the microprocessor businesses and being part of the starting of the PC era, LSI Logic was the system on a chip days and helping Sony come out with the Sony Play Station. And LSI during the early nineties was, I think two years in a row, the best performing stock in the New York Stock Exchange thanks to the Sony Play Station and Sun Microsystems and a couple of other little companies like that. Came over to National in 1996 with an intention of staying here for a couple of years. And as of today, I believe, I'm celebrating my ten year anniversary. So that means it was eighteen years ago that you and I first met. Times fun when you're having flies.
RW: Well why go into engineering? What led you to that?
BH: You know, we ask all of our engineers that because there's, as you know in the United States , there's a shortage of engineers, a shortage of math and science graduates. My first interest was when I got a shrink-wrapped transistor radio that plugged it all together and listened to that little ear piece and, sure enough, there was a voice. Lot of people said they got into engineering because of their first Heath kit or whatever. But - so it was really that first transistor radio. And I must also confess that trying to pick what course of study I would take in college was influenced by a Time magazine that came out while I was a senior in high school. And it had all of the starting salaries of graduates. So I went down the right hand column to the biggest number and then read over and it said electrical engineering. And out of twenty-three pledge brothers that all started in EE that year, I was the only one that made it all the way through. And I'm glad I did.
RW: Took an extra year.
BH: Yeah, took that fifth year.
RW: Yes. So did it take me and this was before the asians arrived in - en masse. It would take a little longer now.
RW: Like forever. Well you - you started out as really a digital guy. And computers and Intel and so on. And - and similarly at LSI Logic. That was essentially all digital except for that screwy English deal where they were going to build analog and they never did. That - but now you've converted over. That - you're kind of an analog.
BH: You know, actually at LSI when we were putting systems on a chip, we had one analog guy in the product group. And this guy was - he was stranger than everybody else. You know, the only thing kind of normal about him was he was his daughter's volleyball coach and so I'd run into him at volleyball games. But he wanted to leave the company and go back to the Midwest . He was the only analog engineer we had. And so we decided that, you know, you go through a crisis like losing a guy who's vital to it that - that time he was working on LVDS and which was, of course, invented at National. And you realize what a crisis it is to lose the analog part of the business. So we had, at that time Wilf had a rec freeze but we had an opening for thirty analog engineers. And then all of a sudden, the CEO Job at National which invented the analog industry comes open and it was a no brainer that - that analog had come to life again, even though the magic of digital killed it for a couple decades. Luckily nobody told National so we retained about seven hundred in the best and brightest analog engineers on the face of the earth. And now we're turning in the results.
RW: Well how about the group that left and founded Linear? They've done well.
BH: They've done extremely well. And Maxim as well. And the other one is ADI which started out in the East Coast. All analog companies, all doing extremely well. It's a relatively low capital business. You don't need state-of-the-art fabs. You know, Rob, if you think back on the nature of this business, I'm sure you've heard Gordon Moore in his speeches say “this is the only industry that has made the price performance gains that we've made, Moore's law, etc. - that if any other industry was as responsive to price performance gains and innovation, today you could buy a Rolls Royce for a quarter, drive it downtown and throw it away”. That would be the equivalent price performance improvements. And if you think about it - if you think about in the history of our industry, we have had this huge capital expense to build a fab, whether - whatever generation it is - as you know 300 millimeter fabs now are upwards of two billion dollars. So the corresponding message that goes to the field sales force is the little tiny things that come out of that fab have got to fill up the fab. So the mantra "Fill the fab". In fact, at Intel, there was a period of time where even Intel which now is one of the only companies - one of the few companies that can get paid for what they do - their mantra used to be "fill the f…ing fab", FTFF. And Gordon Moore, in all of his class, changed that to "fill the facilitized fab". And even Intel was driven by the need to fill the fab. And so when you're in that position and the customer's got multiple choices, there is no such thing as value pricing. And in the analog business and the companies that we just mentioned, Maxim, LTC, ADI and National Semiconductor, you don't need to fill the fabs. I mean, the fabs are fully depreciated for the most part. It's a matter of putting into those fabs products with the right margins, right ASP's and, you know, that's a matter of product definition and education of the sales force, and the customer. And it's a much better business.
RW: Yeah. Or putting it in another way, it's not such a crazy business as the digital world. I wonder if , for example, at LSI Logic and whether we had positive retained earnings over our lifetime. I've never gone back and looked at the numbers.
BH: Yeah, I think a couple of the companies were pretty close. You know, the thing about digital - if you think about this - analog - I mean, Bob Widlar invented the industry back in - thirty-five years ago or something. But this digital thing all came about as the magic of Intel and magic of the microprocessor and the PC and so for two whole decades, the universities all shut down any of their analog curriculum. Only Georgia Tech kept an analog curriculum alive. All the others shut it down. In fact, National donated about a three million dollar lab to Berkeley many, many years ago and it's got a beautiful plaque outside that says the National Semiconductor Analog Fab or Analog Lab - for the first two years they taught a digital course in that lab. So what happens is that there's no Synopsis equivalent synthesis. There's no Cadence floor planning equivalent. There's no CAD tools that help the designers. And so analog design is still considered a black art. So it goes around in the minds of the people - the design engineers. And that is incredibly valuable and takes a lot of the craziness out of the business. In digital, quite frankly, when you've got all these tools and you've got all of these graduates still coming out of the universities that want to be digital designers, you know, they come out, they're small fish in a giant pond. And with analog, there's still a limitation on the number of overall engineers shared by those four companies and then to a lesser extent by companies like Texas Instrument.
RW: Well the, of course, there's the Japanese companies. And they do a lot of consumer work in analog.
BH: Yeah, the Japanese - the Japanese are best serving as a role model for the Chinese because they took such a good run at the industry. And, you know, I came to the SIA in this job, the Semiconductor Industry Association, as this brash, young guy that, you know, was sure I could do it better than the forefathers, the founders of this industry. And the first thing I found out is that these guys were all heroes, Charlie Sporck and Gordon Moore and Wilf Corrigan, that what they did vis-à-vis the Japanese, was truly heroic in stopping them and forcing the anti-dumping laws, the 301 trade legislation. I had a role to play in that as well. I was at Intel at the time and had been a sales training for Hamilton/Avnet it was called in those days. And somehow one of the sales guys got a foil from Hitachi and the foil said don't stop bidding until you've won the business. If Intel undercuts you, go ten percent below them. If they undercut you again, go another ten percent. And it was written - that foil was done by a guy that used to work for me at Intel. And I recognized it right away and so I was the go between. I passed it onto the executives and they decided that - that this was very serious. And that kind of led to all the dumping thing. But the fight that the founders of the Semiconductor Industry Association put up at that time twenty-five years ago was truly, truly heroic.
RW: Which, in many other industries, we didn't do - television -
BH: Television, and the automobile. And, you know, that was - that was really the - the kind of vision is if you want to have happen to the semiconductor industry what happened to the television industry, you know, don't support us. Now having said that, Rob, here's where we are today. The Chinese have learned from the best and the worst of all of these would-be, want-to-be, countries that - that want to get what the United States has. Japan was the first. Korea , Taiwan followed. What's the difference? The difference is Japan is a very tiny set of islands. They ran out of people very quickly. Their cost of wages very quickly exceeded the United States . So they were no longer the low cost manufacturing entity. And, you know, there's other things that China has learned from the success of our industry here in the United States . One of those things, Rob, is stock options. They know they work. Stock options created this industry along with the other - the other kinds of things that we did to - to get the industry going. So in China , the stock option is very lucrative. Strike prices at par value and no capital gains tax on exercise. And what are we doing here in this country? We're eliminating stock options. You know, stock option expensing is just a fancy French word which means eradication over time. And, you know, the other thing that the Chinese have learned is the importance of education of these undergraduate engineers. So they would send them and - in shiploads over to Stanford and Berkeley and MIT to get the best education on the face of the earth. And now in our infinite wisdom, we're sending them home, “visa denied”. You know, with the capital and H-1B visas. And I always - I always say - Andy Grove was always kind of a hero of mine - working for him for fourteen years and a true mentor and if you think about it, Andy came over here as an immigrant - was a refugee, got himself into the position with a lot of hard work and Intel, over the years, has probably - and by the way, I always say that Andy could have been equally successful in the screen door business. But, over the years, he took that company to the point of creation of the microprocessor and the PC business. PC business got us the internet. The internet got us the search engines and the buying engines and how many hundreds of thousands of jobs have been created by a guy, Andy Grove, who today we'd say go home Andy, “visa denied”? And, you know, that's just another one. That's - so the SIA today, our primary role is to spend a lot of time in Washington and try to convince our politicians that we run the risk of losing the technology lead. And I call it a perfect storm of technology neglect. It's not only H-1B visas, stock options - it's substantially reduced funding of the National Science Foundation. DARPA now has an eighteen month time horizon where they need to get payback and so they put less and less of their money in the university system and more and more into the Raytheon's and Lockheed's of the world where they can get a payback. This little entity that got us the transistor back in 1947 that Noyce and Grove and Moore -
RW: Bell Labs.
BH: Bell Labs, you know, very unsuccessful entity. Right? One out of twenty projects is successful. But what were they? The laser and the transistor, surround sound, UNIX. I mean, you name it. And it's been incredibly important. Well Bell Labs had a source of funding from the telephone calls of AT&T at - I think penny per call or something that was such a gift to the rest of the world. Well Bell Labs now is a tiny fraction of itself and so is Stanford Research and so is Xerox PARC. And so the basic research now has to come out of the universities and that means additional funding from the United States government. One time in the history of our country that nobody remembers anymore except a few of us, 1957 where there was no question in anybody's mind that we had fallen behind in the technology race and that was when Sputnik went up. And, you know, there I was in Fort Dodge , Iowa . The whole family came out and all the neighbors were out and we had a blanket on the front lawn and a bowl of popcorn and 7-Up in bottles in those days. And, you know, you could hear the tittering conversation up and down the streets. And one lady says there it goes. And somebody else says no, that's just a shooting star. And - and you could smell the cigar smoke from the guy across the street. Anyway, then Sputnik went over and it was like the whole block went quiet because if you recall, in those days, certainly not in my neighborhood, but in some backyards in Fort Dodge, Iowa, people had bomb shelters because of the Cold War. And here was a satellite going overhead with a payload big enough to carry a missile right over the top of our houses. So Eisenhower had no problem throwing a billion dollars, which was a lot of money in 1957, at the space race. And he also created the National Defense Education Act to spur interest in math and science. Sound familiar doesn't it? Well the space race caused the mainframe race. So mainframes had to get faster and faster. That got them hotter and hotter. And Noyce and Grove and Moore used that transistor to create the DRAM to replace the ferrite core magnets in the back of mainframes. That created the semiconductor industry. And meanwhile, DARPA took all these mainframes now and tried to wire them together and that, with the help of the University of Illinois and Mosaic, became the internet. And Google and E-bay and Yahoo and the dot com companies that are now becoming very successful. And so it was that important realizing we were behind in those days, it was that important for the government to step in and fund some spending in that area. And look at the payback we got as a nation. We're there again. We're on the verge of losing this technology lead and because of this perfect storm of technology neglect.
RW: Yeah. Well Silicon Valley has been a place of innovation for a long time. What do you attribute that success to?
BH: Well certainly these founding industries - I mean, you start with Hewlett-Packard, right, and this whole garage shop approach to changing the world. And, you know, the - the energy that Intel had, but underlying all that and it's incredibly important not to forget the importance of venture capital in really getting us going. And now we've become Silicon Valley . What I would call and I heard Bill Gates use this term we're the IQ magnet of the world still. The really smart people want to be here. This is where it's going to happen. This is where the chances are best that you're going to surround yourself with a whole bunch of other really smart, high IQ people. And so I think we've still got that going for us. And it's one of the reasons, by the way, that when the best and the brightest do come from these other nations to get their shiny new degree or the PhD, that they stay here. And so it's vitally important that we let them stay here and that we do get rid of this cap on H-1B visas. I think it's the IQ magnet of the world still.
RW: Well I was born here. So that's probably the only way I got in.
BH: Yeah. But, I mean, look how - look how many other states and countries have tried to duplicate this thing, never coming close to what's happened here in the Valley. And it's phenomenal.
RW: Well I see you've got a few photographs and so on. I wonder if we could go and take a look at those on your back wall and explain what those are?
BH: Sure. Well inherited the Chairmanship of the SIA this year. And, you know, my primary motivation for the year is that we not lose the technology lead on my watch. That's why I've been spending time in Washington doing what I consider to be one of the most distasteful things you can do which is spend your time with the politicians. This is a picture of Gordon Moore who was absolutely one of the classiest guys in the industry and clearly responsible for so much of what we see around us. And I'm holding up a fish here. This was out on Gordon's boat. We went out to the Farallones, Dave House, myself and a few of the other Intel guys. And so, you know, the boat was rocking and I went flying into Dave House and he threw his pole and everything over the side and it was a four hundred dollar set of equipment. Gordon says no problem, I got more downstairs. And he goes down and on the way back, Gordon cleaned everybody's fish and then took everybody's picture with their fish. And it was like the - the grandfather that everybody wished they'd a had is Gordon Moore, just the coolest guy and classiest guy in the world. So this is a favorite. I thought - hey LuAnn, is there another Nebraska one over there like with me and Eric Crouch, the Heisman Trophy winner? Anyway, this is Frank Solich who is the coach of University of Nebraska , which is I still have season tickets and go back for one game a year. This is another classy guy, the governor of Maine , Angus King, who was really responsible for keeping National Semiconductor and for us putting Fairchild back into Maine . He was just incredibly tech savvy. Here's an interesting guy. Also a major supporter of the industry and shifted gears when he created the fabless semiconductor industry by putting up the first foundry. And that was TSMC.
RW: Morris Chang.
BH: Morris Chang, absolutely. Very vital to the industry. He's kind of low keyed now. I think he's retired from TSMC. Let's see. What else do we have? We've rung the bell several times at the New York Stock Exchange. National Semiconductor's a New York Stock Exchange Company. Here's a eight-year favorite of mine, Wilf Corrigan. Another one of the tough guys, founders of the industry and the Semiconductor Industry Association and one of the guys that fought the hardest to keep this industry on our shores when the Japanese wanted it the other way. And then we have George Bush. This happens to be at an event when George Bush was trying to solve the energy crisis for the United States . And at the time, National was in the information appliance business and we were touting thin clients. And so I was explaining to him how thin clients work and how it saves energy because you have all the energy spent somewhere where there's a server farm and you put the server farm next to a raging river where you can get cheap energy. You don't have to have it here in Silicon Valley . This was - at this - at this event, there was about five companies demonstrating around the room and he understood what I was talking about. He went to the next booth and they did this thing and President Bush goes (whew) right over my head. He says just a second. And he says Bob, and he calls for his - he says this is electric Bob over here. He says come on over electric Bob. And he turns to me, he says his real name's Bill but I didn't want to yell electric bill here in California . Anyway a lot of memories, a lot of tough guys. And I enjoy looking at these pictures. I got another wall full of pictures over here and -
RW: Well we've changed the world, haven't we? I mean, the world we live in now is different than the world when we were kids.
BH: Absolutely. And it - and it's -
RW: Mostly for the best.
BH: And it's changing exponentially better. Now, you know, if you think back on the cycles of the industry, the first one was that DRAM cycle where all the semiconductor companies, there was only a handful of them, threw everything they had at creating DRAM because that was the hot market. Right? That was, you know, replacing all the ferrite core magnets in mainframes. And, of course, we overbuilt and we went over the cliff and that was the first cycle. The industry only got up to five billion sales out in those days but we thought that was a lot of money. And then the PC boom back in the - in the late seventies, early eighties, all the semiconductor companies diverted all their capacity building PC kits and, of course, we overbuilt there too but we only got the industry then up to twenty-five billion. We thought never get any higher. Boom! Next - the next downturn in the second cycle. The third one was the connected PC boom now also with the cell phone. And we got the industry up to two hundred and four billion dollars in sales out in that internet and cell phone. And, of course, we overbuilt. We overbuilt infrastructure chips and everything else. Now we're on the upward slope of the cycle that followed the dot com boom, bubble burst. Different this time. All of those other cycles were caused by one killer application, whether it's mainframe, DRAM, PC kits, the internet and the cell phone. Now if you look at what's driving the industry, it's PDA's, it's MP3 players, it's digital TV's, it's PVR's, it's these things that show up under a Christmas tree or driving demand for industry and guess what? We now have the industry on a run rate basis at two hundred and fifty billion and we still haven't seen the next killer app in the IT space. Microsoft still hasn't released Longhorn. And yet somehow demand is going crazy. It's being driven by this wide diversity of two to four hundred dollar products. So the question I'd ask is how do you stimulate the next downturn when there is no one obvious place to build a pile of inventory? So I'd say it bodes pretty well for our industry and the future looks fairly bright.
RW: All right. Well thank you.
BH: Thank you, Rob.