Interview with Albert Yu
September 15, 2005
Atherton, California

RW: Albert Yu obtained his PhD in Electrical Engineering from Stanford University and began his semiconductor career at Palo Alto 's Fairchild R&D. He was later recruited to Intel by Andy Grove and began his distinguished thirty year association with Intel. In this 2005 interview, he recounts how as Senior Vice President, he oversaw the development of the 386, the 486, many of the Pentiums and the revolutionary Itanium.

RW: Welcome.

AY: Thank you.

RW: Could you tell us a little bit about growing up, your childhood, your family?

AY: Sure. I was born in Shanghai in China and moved with my parents To Taiwan. It was kind of interesting as my father was opening a sales office in Taiwan and we were supposed to go back to Shanghai after a couple years. But we couldn't go back because the communists took over mainland China . So I went to high school in Taiwan . And then my father' job moved again to Hong Kong . So I moved with the family to Hong Kong . And then I came to the United States for undergraduate at Cal Tech in Pasadena and then graduate school at Stanford. So that's sort of the early days of my life.

RW: What business was your father in?

AY: My father worked for a British chemical company called Imperial Chemical Industries. And he was really sales. That's why, you know, he went to Taiwan to set up the Regional Sales Office. He was in charge of it. Then he went to Hong Kong which was the Asia headquarter for his company. So he was in sales of chemicals, the dyeing stuff and so forth.

RW: What made you choose electrical engineering for your studies?

AY: That was totally by accident. When I was in high school, I was really interested in music and western music specifically. So I was able to get some RCA records, 33 RPM, to listen to them on an old record changer and amplifier/speakers. And those were not very good. So I tried to fool around with equipment and tried to make it better. And that got me into electronics. So for a long time, I wanted to be an audio engineer but, of course, I never became an audio engineer. And so I was very interested in the electrical stuff from that point of view. And when I went to Cal Tech, it's really more of a science institution, lot of people interested in math and physics, and electrical engineering was sort of secondary at Cal Tech. And I guess I realized that, after I graduate, I need to get a job. And it seemed that I could get a job better as engineer than a physicist. So that's how I got into it. But the early days I was really interested in music in hi-fidelity, that kind of a thing.

RW: That's exciting. Now what was your first job then out of Stanford?

AY: That was quite interesting because when I got out of school at that time with a PhD, most of the jobs were on the East Coast - the classical places, Bell Labs, IBM, RCA and so forth. And I didn't really want to move from California . There's only one company interviewed me. It turned out to be Fairchild. And the interesting part was when I went to Fairchild, I was interviewing with Herb Kroemer who was doing laser stuff, who later on became a Nobel Prize winner. And he was at Fairchild. And my thesis was in optics. So I went to Fairchild to interview to see if I wanted to join his group. But at the same time, Andy Grove was hiring people for his physics department. And he listened to my thesis and thought it was very interesting and we started talking. I told Andy, I said I'd really rather not be doing optical stuff and do something different because I was interested to learn something new. So he said well why don't you come and work on silicon stuff? So, it was kind of an accident that I got involved with silicon and Fairchild IN 1967. I could have easily gone to Bell Labs or IBM.

RW: So what did you develop at Fairchild, what processes?

AY: At Fairchild, I was working on devices. One of the first devices I work on was schottky barriers. In fact, I was working with Carver Mead. He was a consultant of Fairchild and I knew him at Cal Tech. So we worked on that quite a bit and published some very good papers. Then later on, I worked bipolar transistors and MOS devices. That was fairly early days if you will. So it's mostly device research. And I think soon after I joined Fairchild, Andy Grove, Gordon Moore and Bob Noyce left for to found Intel. That was in 1968. So I was still at Fairchild. As time goes on, I got increasingly frustrated because, Fairchild R&D lab was in Palo Alto and the manufacturing in Mountain View and the two never seemed to talk to each other. And I felt I developed a whole bunch of interesting devices but nobody ever wanted to pick it up. What's interesting was that Fairchild had a diode plant in San Rafael and the fellow there Ed Browder was interested in what I was doing. So I actually transferred the process from Palo Alto to San Rafael all by myself. I went there every week to get the product into production. That was the most exciting thing for me to get products into production. So I was pretty frustrated after about four and a half years at Fairchild. Andy Grove called me and said, do you want to join Intel? They're looking for somebody in the manufacturing area. I had no experience in the manufacturing area. I was then more of a device researcher. But I thought that's great because I could work on something that actually would go to the customers and go to the marketplace. So I said I'd be happy to come and work on manufacturing. And that was 1972. So I joined Intel in 1972. Intel was in Mountain View , in one small building and everybody was there. It was just a small place and construction was going on to improve the FAB capabilities all the time. I was responsible for taking two inch wafers to three inch wafers in Fab. That was my very first project. And what struck me was how open and helpful everybody was at Intel versus Fairchild where there was a big gap between R&D and manufacturing. And I didn't really know much about manufacturing but lot the people there were very knowledgeable taught me, tell me what the right thing to do and some of this secrets, of some of these processing. Our team also came up with new processes. It was a phenomenal time for me, that I learned and contributed so much about manufacturing. So that was the beginning of my Intel days.

RW: Well that followed Andy Grove too because -

AY: Right.

RW: - he - he had no manufacturing experience -

AY: That's true.

RW: - until he got there.

AY: Right. And I think that he hired two people, one responsible for fab, the other responsible for assembly tests. So he hired people that knew those areas that reported to him and he learned from them. And I reported to Gene Flath at the time who was responsible for the fabs.

RW: So what else did you do at Intel?

AY: I was in fab for a couple years. And then I think it was a big economic downturn in '73, '74 timeframe and lot of our product ran into trouble in reliability, moisture and stuff like that. So they asked me to go into reliability engineering. And I was not very happy about doing that. I thought, you know, manufacturing our products is where I wanted to be. Reliability seemed to be something that didn't require that much technical knowledge. It turns out I was totally wrong. The technical content of reliability engineering was very high, plus I think that was the first time I began to interface with customers. Before I was the inside person. I dealt with R&D, I dealt with manufacturing. But in reliability, you got to talk to the customers because there were problems. You got to go understand the problems, work with them to solve it. So that turned out to be one of the job I learned the most, both on the technical side as well as the interfacing, talking to customers and understand what their concerns and how to make things better for them. So that was very exciting. And then after that, I became head of Technology Development. Actually Technology Development was the heart of Intel if you will. I was developing EPROMS and the first fast static RAM. Those were really very-very successful products. As you know, Intel was in memory at that time beginning with 1103, a dynamic RAM. That's when I did the two inch to three inch wafer conversion with 1103. And then later on we developed EPROM and the SRAM. So that's what I was doing at Intel. That was until about '77. So it was about five years from '72 to '77.

RW: Could you - thinking back to those days of the two inch and three inch wafers, could you ever envision the size they are now? I mean, if somebody had told you that, what would you have said?

AY: The scale has grown so much, it's unbelievable. Compared with today, running twelve inch wafers with hundreds of people in the fab. When I did the two inch to three inch conversion, I had only three people working for me - two engineers and one technician. The four of us basically did the whole process. Now, it's an army of people doing that. Of course, everything's much more complex. The geometry is much smaller and everything else. I had no concept, at that time how expensive everything would become like today. It was only three, four people on a project at that time.

RW: So how long did you stay then at Intel, that first time?

AY: I stayed until about '77 and I was working on memories and technology development. I saw microprocessor being done as Intel introduced the first microprocessor 4004 in 1971. And I get to know Federico Faggin, the 4004 designer very well and talked with him quite a bit. I got very intrigued by that. And then something kind of clicked in my head in 1976 that we already knew about the Moore's law, that transistor density would double every eighteen months or something of that sort. So if you did a simple extrapolation, you said Oh my God, we're going to make millions of these microprocessors. We were making lots of memory chips, and the application was obvious. You put them in computers as the memory usage in the computer was limitless. But for microprocessor, it is a different picture. Where would you use it for? At that time, they were used for traffic controllers and some industrial controls, pretty mundane application, not particularly of high volume at that time. In 1976, I sort of came to the realization that there's going to be millions of microprocessors produced as time went on. So where was it going to go? Well I thought it was obvious: It's going to go into the home. So home computers were going to come and be in the millions. But remember that's 1976. That's before Apple was founded. I actually got so excited about that and left Intel and got together with a couple of friends and started a company doing home computer. That was really early days, if you will. There's no floppy disk drive. There's no operating system. And, in fact, only thing that's available is BASIC from Microsoft. And so we worked with Bill Gates about having a BASIC running on our machine and developed a whole bunch of software. But we were way ahead of its time. Looking back, home computer didn't really become real until like 1994 or 1995 when the CD ROM came along and Pentium processor came along and hard disc and all that stuff. So I was little crazy and way ahead of its time and did a home computer called Video Brain. And, it was kind of interesting for me because I had to sell the computer through retail channels, Sears Roebuck, Macy's, all these people and I went on the retail counters to sell to the consumers directly. It became very clear to me after a little while that people didn't know what that was. And at that time, about '78, '79, Apple introduced the Apple II computer. That was aimed for a totally different audience. It was for computer enthusiasts and so forth. And the Apple obviously took off from that point on. My home computer never really took off. But I learned so much about it. Would I do that over again given knowing what I know? I think I probably would have done it over again because I learned so much during that period of time on everything - not manufacturing, not R&D but marketing, sales, cash flow, products, getting software put together on the computer that helped me a lot later on. So it was a phenomenal experience but not financially very successful.

RW: Well who else was in there with you at Video Brain?

AY: The couple of Intel people who came with me - the technical guy with me was from Fairchild - was not from Intel. The two other guys left Intel, went with me, Bruce McKay in manufacturing and Rich Melmon in marketing that went with me. I invited Andy Grove to take a look at our system and he was skeptical and said well why anybody would need one. And he was skeptical about Apple also. That was such early days in 1977- '78 timeframe.

RW: Now did - did David Chung - did he -

AY: Yes David Chung was a technical guy that worked with me from Fairchild. Right.

RW: And you - you used the F8?

AY: That's correct. We used the F8

RW: Why did you choose the F8?

AY: Well it's because that was the chip David designed. We should have picked the Intel chip but we didn't. We didn't quite understand at the time the importance of compatibility software and things like that. It was not clear at the time. There was no operating system also at that time. CPM wasn't even quite there yet. So that was before "microcomputer" came along. PC came later but with Apple and the CPM Machines. Rich Melmon was our marketing guy from Intel

RW: Yeah, he recruited one of my guys.

AY: Is that right?

RW: Yeah. Mike Peak .

AY: Oh sure. Of course.

RW: and then about a year late, Mike came back to the…

AY: Right. That's good. I remember Mike really well. He's a terrific guy.

RW: He really is.

AY: Yeah.

RW: So you went back to Intel.

AY: Well not quite. Video Brain was not successful. So in 1979, I became a pioneer again. I decided that home computer was not what people want at the time. However, there was an opportunity in China that was before the relationship with the US got normalized. They were in such a poor state that I talked to some people there and they're very interested to learn about microcomputers because they used Russian mainframe that broke down all the time. So we decided to sell microcomputers in China with CPM operating systems. It was really distributing- as we didn't build the computers. We bought the computers from companies like Dynabyte and Cromemco and sold the computers, totally with software to China . We also trained them how to do programming in FORTRAN and BASIC and all that stuff. So we did that for about a couple years. Again, that was very, very early on. China at that time, was totally different from China today, where the business is just booming now. At that time, they were really a closed communist country. They really didn't understand profit and things of that kind. So it's very difficult actually. But again, I learned a lot during that period of time and I went to China about four or five times a year from 1979 all the way to 1981. That was quite successful and we made some money out of it. But I felt it's not something long term for me because the environment-I was not pro-business. So in 1981, Andy Grove called me and said, “you had your fun now. Maybe it's time to come back". So I did. So I was gone for four years from Intel.

RW: So what did you do when you got back?

AY: At that time in 1981, the major problem was Japanese companies were assaulting Intel's memory business. And the biggest thing I remembered at the time was a fellow Anderson, at HP saying that America 's DRAM quality was much, much worse than that of the Japanese. And so I was put in charge of Quality and Reliability. And that was a major challenge how to get better in quality, not only in reality but also in perception. Very often quality is perception as well. So I did that for about three years and we were able to improve the quality of all the Intel products by a factor of thirty. The defect level got much, much lower. It was something that Intel had to do it and I did it. I feel really proud of it. That was kind of a broadening experience; I became more of a senior manager. Before that, I was really a first line manager. Even at Video Brain, which was a small company, that did not have much of a management structure. But during those times, I learned to be the second and third level manager. And that was a phenomenal experience. Learning management in Intel was probably just about the best place there is to learn and to make mistakes with people help you and grew. And so that was a very good time for me. But eventually, as I got so excited all the way in 1976 about microprocessors and computers, I wanted to go back to microprocessors and work on computers again. And so that's how in 1984, I got together with Dave House who was in charge of the microprocessor group. And I said I really wanted to work in the microprocessor area. And he said sure come on over. So I started in the microprocessor group in 1984 and started out doing product planning and very quickly become Assistant General Manager with Dave. And as time went on, I become “two in the box” with Dave House for a long time. I think we were working together for about seven years. And that was when 386 came out. That was a huge success for Intel. The design of 386 was doing so well it taped out right after the July 4th weekend because nobody else were using the computers. So the computer's available for design. So we were ahead of schedule by several months. And the 386 came out and Compaq jumped on it and product became a phenomenal success. It was 1987 when Intel really hit big time in terms of profit and revenue and everything else. And I was right in the middle of it; I was so happy. Then, of course, after that the a whole bunch of processors that I was responsible for - 486, Pentiums and so on and so forth. So I'd been responsible for microprocessors since basically '84 all the way until year 2000. So that's a long time, like sixteen years. And I worked closely with Dave House, Paul Otellini and Carl Everett and whole bunch of people at Intel. I thought I had the best job in the world.

It was appropriate to add a note here that when the home computer finally emerged in high volume in the '94-'95 time frame with the then power Pentium processors, good graphics and graphical operating system-Windows, new CD-ROM and enough software applications that appealed to home users, it is amazing that our original concept of a home computer in '77 was very similar to what eventual became the real home computer, 17 years earlier. I was happy to be a key player supplying the “Intel Inside” of the home computers.

RW: How about the attempts to obsolete the X86? First there was a 432 and then the - what is it the 860 -

AY: 860, right.

RW: -and then there's now the Itanium. And pretty much these have failed in these attempts.

AY: I think Itanium has not failed. I think the Itanium chapter has not been written yet. My understanding, as I retired from Intel three years ago in end of 2002, it's gaining momentum but not to the volume as the Intel architecture yet. But let me talk about that for a little bit. I think it took Intel a while to realize the power of the franchise of the instruction set that's totally compatible. And IBM did that in their 360 days and I think took Intel a little bit to understand it. And there was always technical ideas coming along. If you do something different, you could have a higher performance, a better processor. 432 was one of those that was supposed to be much better. I was not directly involved with it but I was very close to the team. And that was supposed to be redundant and reliable, all these good features. But it was too complicated and didn't have the performance. And then was the RISC challenge. That was a big battle where Sun was the first one with Sparc and MIPS with their chips and then IBM Power PC. And they were saying the RISC architecture was better than the Intel architecture and they're going to take over the world and so on and so forth. About that period of time, we had two products going on. One was the 486, which was totally instruction compatible with 386. Therefore, all the previous software could run on it. At the same time, there were some technical virtues of RISC and in fact, many of these features were incorporated in the 486, which outperformed some of RISC chips So we decided to do another chip which is not to replace 486 but like a coprocessor if you will. That was the 860, which was a very much a coprocessor, focused on floating point and so forth. And again, the volume was too small. The whole economy of the standard product is volume. And the 386, 486, Pentium – the volume is enormous. So the costs came down and although the chip may be bigger than if you design something specific for the application but the cost of that would be much lower for a high volume processor. A general purpose processor tends to be lower cost than anything that you design specifically. The 860 was successful in the graphics processing area, but the volume was small. The Itanium is a different story, as the idea of going to 64 bits is a very different story. We believed at the time that 16 to 32 bits transition was going to happen by definition because all the mainframes were 32 bits. But 32 to 64 probably would take some time. And that transition takes more than technology. We really need a good partner to make it happen. HP emerged as the leading partner because they have developed this technology that looked very, very good. We signed a deal with them in 1994. Dave House and I were involved with Lou Platt and sorry to hear Lou passed away. He's a really terrific guy. And so we combined a new instruction 64 bit computer but still had built into it compatibility of 32 bit software. So it would be best of both worlds plus HP would be a major computer user who would use the 64 bit computing. So it's really a combination of different things. I didn't think we ever thought about obsoleting the Intel Architecture because we felt that as the market got bigger, the Intel Architecture would continue to go on to the highest volume. But we need something higher end for 64 bit and for high end servers. So that was what it was designed for. I think up to this date, the volume is still relatively small as compared with the main Intel Architecture. So that's in a few minutes, the history of all that stuff.

RW: Well the 64 bit X86, was that done an official Intel effort or was it done in the back room?

AY: It was done in the back room. It just like the 386, you know, where it was done in the back room also. But it was pretty clear. We did not announce that to the outside world. We put in the 64 bit into the Intel Architecture. We put in some other features we didn't typically talk about anyway. But we put it in but don't really announce it until we need to. So that's what we did.

RW: But was there some sense the Itanium was a competitor. Is that one of the reasons it was held back or -

AY: In a way it could be a competitor if we were not clear on the positioning. Itanium was never intended to be a replacement for Intel Architecture. It was never thought of that way. It's always to be for high end servers. And I think there probably some confusion. Some people might think that we're going to take Itanium to replace Intel architecture. That's never been the intention. It will be an additional architecture, and that was the intent. And the intent was also that you could run the 32 bit software without any change. So it had two instruction sets: Intel X86 instruction set as well as the 64 bit instruction set.

RW: How does on these development - how does the Israeli design - they've done several of the Pentiums, have they not?

AY: Right, right.

RW: What - what gives them their edge?

AY: Well the Israel design center started long, long time ago. It's like 1975. They're right next to Technion which is one of the best universities there. So Intel is physically located next door and has been able to attract the best brains out of the university. And so basically the best brains of Israel . So Intel have had a tremendous advantage in that space. Now it took a long time for a team to learn how to do a processor well. It's non- trivial and they had some very successful products on some Pentium variations and also they had some problems with some products that we have to discontinue. But the most successful product which now is biggest of all is Pentium M. Pentium M was a product that was actually got started just before I left the microprocessor group. At the time we said gee, frequency kept going higher, the heat kept on going higher too. You can't go that way. And particularly if you think about anything to do with notebook and something like that. You want the performance but you can't have the heat. So the Israel team came out with some really novel ideas that took the Pentium III architecture and improved it such that the performance was just as good or comparable to Pentium IV but the clock rate was less. The clock was directly proportionate to heat so it turned out to be comparable to performance than Pentium IV but much lower power. And that became a huge success. Not only the Pentium M processor but the combination with Wi-Fi, as Centrino is branding of the processor along with the Wi-Fi THAT become a huge success in the notebook area. I'm out of Intel now but it's very clear going forward the dual processors, quad core processors and so forth going to be based on that Pentium M core which is much lower power than the Pentium IV. So Israel had done phenomenally but it took a long time. It started in 1975 and it took them like twenty-five years to get to where they can do world class designs. It's a phenomenal team.

RW: Well, what do you attribute the success of Intel to overall as a corporation? I mean, they managed so well. And typically promote from within. The - they're not bringing in the people from the outside and like you see so often in American corporations when they need somebody to save the corporation.

AY: Well, that's a very good question. I think there are several factors. One of the factors was the people at the top from the beginning. You know, Gordon, Bob and Andy each had their strengths but they each also had their weaknesses. But the three of them in combination was almost perfect. Andy, in the early days, was a chief operating guy. Gordon was the visionary and Bob was Mr. Outside. And there's very few people can do all three things well. And they worked so cohesively. I've never seen anything like that, in all my years, not only in semiconductor industry but in other industries. And then I think the transition of the leadership has been extremely smooth. Bob passed away and Gordon became Chairman and Andy became CEO. Recently, Craig Barrett became Chairman and Paul Otellini became CEO. But the transition was very smooth. This happened with the transition of leadership of the microprocessor group as well from Dave House to me to others. It's like nothing has changed. So there's no major disruption from a culture and from a leadership point of view. There is one thing that I get so excited about Intel is that I have worked with just about the smartest, the best people I ever worked with. And that is why I went to work every day. It's because it's so fun to work in that environment. The combination of the best people I ever interacted with, the leadership just seemed so consistent and the transitions were so smooth. For every company, the issue is always people. And I think that Intel has, for a long period of time, had some of the strongest people around. And very typical if some people left, it didn't matter because the bench was so strong. People moved up and things went on. I think this is the number one strength of Intel. The other part of Intel that's really strong was, I think that very few companies can do that, was an incredibly strong product line. You need a phenomenal manufacturing operation and you need a phenomenal marketing and sales team. Very few companies have all of them so strong. Intel always was the leader in silicon technology and the product line always has been very strong. And Andy Grove - I credited him for getting the marketing absolutely to the highest level of the hi- tech companies. I still consider Apple doing a very good job in marketing but Intel is probably equal. The sales and marketing teams were really strong. I remember going out in the field all the time with the sales people. Those guys were really knowledgeable about products and have a good relationship with the customers. So if you have all these together, everything shines. Now sometimes not every one of them were at the peak. Sometimes the products were a little less than what it ought to be. Sometimes the manufacturing was not cranking out as many as they should, or some of the marketing programs were not as coherent as they should be. Then you were not as good as you can be. But when all these come together, it's unbelievable. And Intel have a long period of time where all these products, manufacturing, sales, marketing all came together. And that's very rare among companies. I'm currently involved with many small companies and it's really difficult to have all these pieces with excellence all together.

RW: Well what part do you think the Intel University internal training plays in that?

AY: Very, very big role. Intel University offers technical as well as management classes. The technical classes are not unique. You can go to universities and take technical class. But the management classes were really, really good. There are classes about Intel culture; how to do performance management etc. and that's very uniquely Intel. You can't learn that anywhere else. And these were taught mostly by Intel people. I was heavily involved with teaching when I was at Intel and it's phenomenally satisfying because every time I'd teach, I learned something new. And it was always great interacting with people who have zillions of questions that were just really stimulating.

RW: Guess it's really a shame because they - you wouldn't take some MBA guy and say congratulations, you're really doing well. We're going to make you an engineer. But they take an engineer and with - with really no training say, okay, now you're a manager. And how do you learn? And that's what most of us gain.

AY: Absolutely.

RW: We just stumbled onto it. And it was wonderful to be told there are three kind of meetings. And here's the rules. At other companies they don't know. They don't know that.

AY: That's one of the most interesting things. I've been at Intel so long, almost thirty years. And now I am working with a lot of small companies, trying to help them grow. I sit on their boards. I have to tell them a few things you have to do like manage by objective. You have to have a planning process and something simple like a waterfall charts showing how your forecast vs. actual every month. These are intuitive to some of us but a lot of people have no clue. And so I've been trying to help lot of these companies and to instill some of these disciplines in managing. I think Intel's really in a class by itself. And I think a lot had to do with Andy himself. From very early on, he taught the middle management people himself. And later on, he taught a class at Stanford and Intel on strategy. And that's a different level of stuff but he's been a phenomenal teacher. I think many, many people from Intel benefited from the teaching of him. So his influence is just unparalleled.

RW: So you're out now. Are you a venture capitalist or what do you do now?

AY: I'm not a venture capitalist in the sense that I have couple hundred million dollars to invest. That's not what I do. I work with a number of venture capital firms in looking at different investments and if interesting to me, I would put in some of my own money and sit on the boards and things like that. I sit on seven boards right now, three of them are public and four of them are private. And it's really a blast because, like I say, I learned and practiced so much at Intel and I feel I can really benefit these people. And also I'm learning new things. As I mentioned, I've been really in the microprocessor space for a very long time. None of the companies I am involved with are in microprocessors for obvious reasons. I got bored with it. I got to do something different. So I'm learning new things. So that's truly, truly exciting. And the other thing I am doing is non-profit. I have a foundation that supports medical research, biotechnology research, stem cell research at Stanford , Cal Tech and UCLA. And that's lot of fun. I'm learning new things. I interact with some of faculties, students, some of the smartest people around creating new stuff in biotechnology and so forth. I don't know much about it yet. I'm still in the learning process. And it's great fun.

RW: Why is this Silicon Valley area - why is it so innovative and what are the factors that really seem to make it work?

AY: Well I think a lot to do with how do the people with different backgrounds interact: technical people, marketing people, venture capitalists.... You run into people, you're your neighborhood, or in restaurants, with different disciplines and you talk to each other and sparks fly. That's how it goes, right. And we have such a free environment. I'm an immigrant to this country. I never lived outside of California in the US . Lot of people asked me was there any discrimination as I'm not a native born American. I never felt that. I felt I'm just like anybody else. And here you have people from India , from China , from everywhere. So it's really a phenomenal place that you have all these people from different area, different views and the sparks fly and creativity come about when these people interact. I think it's a unique place.

RW: Well you're also allowed to fail.

AY: Absolutely. I've failed many times. It was fine. I think that's another important point. Failure is okay if you learn from it.

RW: Well, in Japan , it's not an option.

AY: In China , it's not so good either. And it's not so good in Europe . People look at you Oh, you failed. Here you fail, you move on and you learn a lot of things. A lot of CEO's that we recruit failed before. But he learned from the failures and say "Oh my God, I'm not going to do these and those again." So in some ways, it's a blessing. So absolutely, you're right. We're not afraid to fail. Learn and move on.

RW: Well thank you so much. Been very interesting.

AY: Great. I want to show you a couple of thing if it's okay.

RW: Yes.

AY: The first one is a plaque that I got when I left the microprocessor group in year 2000. It showed eight chips that I was responsible for from the early 386 all the way to Pentium IV. And if you see the top line reads "1.3 billion served". It was like McDonald, we served so many microprocessors. I'm really proud of that obviously. And the other thing was that I've written two books. I hope you can see them. One of them is in Chinese on the right and it was a best seller in Taiwan and China . It was very interesting because Intel was perceived to be a big, aggressive company in Taiwan and China . And my book humanized Intel to a large extent. They said Oh, you know, these Intel people were hard working people anyway. So it had some pretty positive effects on Intel's image in China and Taiwan . And the others one's a English book on "Creating the Digital Future" to talk about my thirty years at Intel. So anyway, these are the things I just thought I'd brought and show it to you and the audience.

RW: Great. Well thanks again Albert.

AY: Sure. It's a pleasure.