 |
 |
 |
Gartner Fall ITXPO Symposium 2002 Craig Barrett MARTIN REYNOLDS: Good morning, everybody. We're fortunate to have with us this morning the chief executive officer of the company that drives the technology in the PC industry. I'm going to be interviewing him with Jackie Fenn, vice president, research fellow, Gartner. I'd like to invite you to welcome me in joining Craig Barrett, chief executive officer of Intel Corporation. (Applause.) JACKIE FENN: Well, Craig, we have some interesting questions. One of the things is around Intel's success to date. By many measures, Intel is highly successful, you have a dominant position in the industry, you're highly profitable, and somehow you've avoided the big, bad, giant reputation that others in the IT industry have played with at times. What we want to do today is focus in on some of the secrets of Intel's success and particularly how you're going to maintain that going forward through some of your technology investments, and particularly in the light of the economy concerns that we heard about. So what would you say is the hallmark of the Intel brand that's got you where you are today? CRAIG BARRETT: Well, I think basically a brand is simply an extension of your company. And our company stands for moving technology forward. It stands for the reliability, the verification, the support you put behind your technology, your ability to communicate that to your customer base, your ability to deliver high-quality products to your customer base. You know, I sometimes harken back to the big learning we had about brand, customer service, and how you stand behind your brand. Back in 1995 or so, some of you may remember we introduced the first Pentium processor and it didn't quite multiply and divide all 32 bit numbers accurately. It did pretty well until about the ninth decimal point, but some of you were worried beyond that. We took the classic engineering attitude of who cares what happens beyond the ninth decimal point. Most calculations don't go there. So it wasn't a practical issue; it was only an issue of engineering significance. And that was absolutely the wrong attitude to take. If you have a brand, a product, if your customer believes in your product, you stand behind your product. Then if your product doesn't work as advertised, you do something about it. In that case, we quickly realized we had to replace those products. And that leads forward to today where I think we're still probably leading the way in terms of telling our customers all the errata, the specification changes, in a real time basis associated with our products. So there are no surprises. The customer has faith in your verification, what you say your product will do. They have faith in your road map, the direction that you'll stand behind the high quality product that you put out. So we just group all of our activities behind that general direction, that general push. MARTIN REYNOLDS: So this is the secret to satisfying the customers. But what about the business objectives of successfully running such a demanding business? We heard Carly Fiorina talk about investing $900 million in printers this morning. What kind of investments do you have to make to keep the processors running forward? CRAIG BARRETT: $900 million would last us about 45 days. That's not demeaning to Carly because in that printer business, $900 million is a lot of money. Our business, $900 million is not a lot of money. Last year and this year, our capital budgets combined will be somewhere in the $12 to $13 billion range. So we're doing $500 million a month, basically. But it's a capital intensive business. If you want a leadership position from a technology standpoint, from a manufacturing standpoint, and we believe that means from a product standpoint, then you have to invest to stay at the leadership edge. And that's just an expensive proposition in the semiconductor business. MARTIN REYNOLDS: So again, every year you're betting $5 or $6 billion on the market for the following year? CRAIG BARRETT: Well, we've been betting for the future of the business for the last 20 years, basically. You build these new factories in anticipation of products that haven't been designed yet, in anticipation of design wins and customer demand that hasn't happened yet. So you're always betting one to two to three years out. You have to have a basic belief that the technology is going to continue to advance, that you'll have leadership products. And we've been betting that way for 20 years, and I think by most objective estimates, it's been a pretty solid bet so far. We're doing it today. MARTIN REYNOLDS: Anything that would make you decide not to place that bet? CRAIG BARRETT: No, the unique thing about technology is it does not recognize economic recessions. Moore's Law has been around for 35 years, and good for another 15 years or so with the silicon technology we have today. We drive ourselves to really be the steward of Moore's Law, to make it happen. And I'm absolutely convinced we'll see that doubling of processing power, doubling of memory density, doubling of the integration of functions with integrated circuits for the next 15 or 20 years. JACKIE FENN: Now, the technology and the engineering might weather the economic recession, but how about the sales power? The ability to get that into the market? CRAIG BARRETT: I'm a firm believer that technology doesn't slow down, it's not going to slow down in the marketplace. There's going to be a continuing demand for new and greater capability, whether you look at processors or memory or communication chips. You can just decide what you want to provide in better processing power, security, Hyper-Threading Technology or dual multi-thread processing, whether you're talking about different, faster memory devices, whether you're talking about wireless communication being integrated into the chipset or into the processor. All of those technology advances will continue to happen. The applications will be there, the customer demand will be there. I don't want to say that you could extrapolate forever, but I don't see any slowing down in the near term or in the next five to 10 year period in the sort of technology that we can bring to the marketplace, and that customers, such as you have here today, will find it useful. MARTIN REYNOLDS: One of the things we've seen in the press is that you've been quoted as seeing the economy turning around. Now, of course, the press will jump on anything these days, if they see such mention. CRAIG BARRETT: I've noticed that, actually. MARTIN REYNOLDS: Tell us what you think. CRAIG BARRETT: You know, I woke up this morning and they were interviewing someone on CNBC who is a supplier to the semiconductor industry, and this individual's stock had been hammered yesterday, and the first question they threw at him was, "Craig Barrett said yesterday in Seville, Spain, that the economy was going to recover in 2003." MARTIN REYNOLDS: You thought Seville, Spain was safe, right? CRAIG BARRETT: Well, my first thought was who is Craig Barrett? My second thought was __ (Laughter.) CRAIG BARRETT: __ why would anyone say that? (Laughter.) CRAIG BARRETT: Let me reiterate what I said because the press is kind of hyper on any economic issues today. When asked when I thought there would be a recovery, I just said, well, let's look at the four issues that have caused the current slowdown. You have a synchronous slowdown or recession amongst three major economies: Japan, Western Europe and the United States. We had a Y2K overhang; a lot of buying in 1999 of IT equipment in anticipation of the Y2K bug. We have an over-investment in the communication sector, substantial over-investment. Over-investment, whether it's long haul capacity, all the fiber in the ground, wireless equipment, but a huge over-investment. And lastly, we had the dotcom meltdown. All those happened synchronously in the last couple of years. They're not going to all last. The excess capacity will get worn off. The Y2K issue, you know, all the computers bought in 1999 are sub-500 MHz. If anyone in the audience or you guys really believe that a 500 MHz machine will give you the same sort of employee productivity, capability, security aspects, whatever, as a brand new 2.5 to 3 GHz machine today, then the conversation is over. I'll just go backstage and you guys will entertain the audience. (Laughter.) CRAIG BARRETT: But those four things have to end, you know. I do think the economic recovery of the high tech sector will be a lagging indicator. Corporate profitability will happen first. That will be driven by consumer spending. When corporate profitability happens, that will translate into IT investment again in the corporate sector, which is really where the slowdown is. So I look at all that, and everybody kind of thought it would happen the second half of this year. MARTIN REYNOLDS: We had help from some large companies here, right? They sort of forgot some funds that they squirreled away as revenue that really weren't. Caused a sudden crisis of confidence. CRAIG BARRETT: But I'm just optimistic about the future. And I did say in Spain that I think the communications sector will recover slower than the computing sector. I think you had John Chambers here yesterday who kind of agreed with that. It's that overcapacity in the communication sector is just going to take a little bit longer to work off. I think the computer sector is relatively ripe for further investment, further capability, further productivity brought to the enterprise segment, and a lot of excitement brought to the consumer segment. MARTIN REYNOLDS: Let's just talk about productivity for a moment. What do you see for our audience that they could invest in that would boost productivity? CRAIG BARRETT: There's an immense amount of technology being brought to the marketplace and not every piece of it is good for every company. I think you have to choose judiciously what fits your company model. If you look at a company like Intel, not that we're typical or atypical, but we are investing in productivity everywhere from making our shop floor more productive, we are a big manufacturer and the instantaneous access to information, analysis, feedback loops, decision making is absolutely critical. Processing power and software make that happen. If you look at how we sell our product to our customers, we sell about $25 billion worth of product over the Internet, 24 by 7. MARTIN REYNOLDS: Bigger than Amazon, then. CRAIG BARRETT: 24 by 7. Essentially no human involvement. Let your salespeople be more productive, makes your customers happier, gives you better feedback into how to load the factory, and that's back to the first part of factory control. You want to be loading the factory with what people want to buy. If you look at the other side of our business, we buy nearly $10 billion worth of raw materials and capital equipment each year. We essentially do all that over the Internet. If you look at the other sector of our business, interfacing with our employees. All of our employee benefits, everything is online. So those are the things peculiar to Intel. Your audience should look at their business, not just from an IT perspective but how does IT make your business more productive, how does it make it more competitive, more efficient, more profitable. And so your IT people work with your line of business people, and you make the investments where it makes sense for your company. There's no general rule, no general dictate: Do this because Intel did it, do this because IBM says you should do it. It's really what makes sense for your company. JACKIE FENN: Now, Intel also has some innovations in your own IT department that may strike at the core of a lot of folks here. Could you tell us about innovations like the split CIO role and some of those? CRAIG BARRETT: Well, we decided some time ago that there's a natural tension between investments in IT infrastructure and the demands or efficiencies that your business units require. So we split our IT organization really into two halves: the classic IT infrastructure part, and then the individual who is in charge of IT applications or Internet applications for our different lines of businesses. So we really have an e-Business person and an IT person. They're married at the hip. They are the co-IT managers or directors for Intel. That way we continually work the tension that exists between those two, the tensions between I don't want to spend any more money for IT because times are bad, and I need to spend more money to make my business more competitive. JACKIE FENN: What are some of the Intel innovations coming out of that around Web services or other emerging areas? CRAIG BARRETT: One of the innovations is the ability to take the sales information and basically go back to our factory network and realign our factory network almost overnight. We can't do it almost overnight. It takes about three days, still, to do that, but with the old paper and pencil routine, it was taking us effectively 45 days to reload our factory in response to the instantaneous demand coming in the front door. For a company as large as ours, that's a huge potential inventory savings. Our inventory atrophies a little bit faster than ripe bananas typically. (Laughter.) CRAIG BARRETT: Come on. How many of you have bought a brown banana recently? It should be nice and yellow and firm, and we want our products to be exactly the same way. And they're not that way if they sit in inventory for two, three, four months. So you want what's coming out the factory's back door to be exactly in tune with what the customer demand is. MARTIN REYNOLDS: So when was it you had a 45 day process and you got to the 3 day process? How long ago did that transition happen? CRAIG BARRETT: We basically, over the last two years, have been able to transform the system from a very manual, spreadsheet intensive, paper, people intensive process to a much more automated system. MARTIN REYNOLDS: So this is a relatively recent shift. CRAIG BARRETT: It's been an investment over the last couple of years. And by the way, it's continuing because three days or four days or five days is not good enough. I'd like to be able to do this basically in a 24 hour time period. MARTIN REYNOLDS: So this is actually a key example of what Gartner described as the real time enterprise: Cutting cycle time for business advantage. Can you identify how much you think you've invested to get to that savings? CRAIG BARRETT: Well, if you look at our entire IT investment, we're in the greater than $500 million a year number in terms of both hardware and software, application support side of that. I can't pull out that exact one, but we are one of the biggest SAP implementers in the world, and it's really SAP that is the core of the factory loading, planning, customer service type of the network. MARTIN REYNOLDS: But the savings, of course, are enormous compared to the investment. CRAIG BARRETT: The savings opportunities are enormous. And they're not always the easiest to calculate in the world because when you look at inventory savings, these are really kind of opportunity savings or these are savings of product that I didn't make and didn't write off. So it's not instantaneous efficiency, ROI back into the system. But if you can get better customer service and avoid those inventory write-offs, it's a huge benefit to the bottom line. MARTIN REYNOLDS: So this is an example of where Intel has real benefit, then, from IT. And there must be parallels for many of our audience out there. CRAIG BARRETT: I would sure hope so. JACKIE FENN: Earlier you mentioned the 3 GHz PCs coming along. We did a little audience survey earlier and we asked the question what do you really want from your PC going forward? So perhaps we could roll that video. It's a user focus group for you. (Video plays and ends.) JACKIE FENN: There was a theme there around the integration of particularly the networking and radio kinds of capabilities. I know that's something that you're focused on. Tell us a little about that. CRAIG BARRETT: I absolutely think the wireless connectivity and the combination of both the wireless WAN, wide area network, which is GPRS or 3G, and wireless LAN, which is any of the 802.11 varieties, a, b, g, x, you pick your protocol, but building that capability into PCs, in both laptop and desktop PCs, I think is key for the future. We're coming out with our computer OEMs early next year with basically built-in dual band 802.11a and b capability as a standard feature in laptop PCs in anticipation of all the hot spots that are being built up at conferences like this, but at airports, restaurants, hotels, et cetera. You don't have to think very hard to figure out you'd rather download in a hotel at 10 or 50 meg than a dial up modem that's somewhere between 25 and 50 K. I just came back from a European trip, and I was getting up in the morning and downloading e-mail every day, and it would take anywhere from 15 to 45 minutes to do that on a dial up modem. I could have done that sub one minute if they had an 802.11 connection, those hotels. We think from a business perspective it's a huge opportunity to put excitement and productivity and capability and efficiency back in the PC. And all you have to do is think a little bit further to, well, all right if you do it at the laptop, maybe you ought to do it at the desktop. If you do it at the desktop, that also allows wireless connectivity at the fixed desktop in business but it also allows it at the home. And you start to look at the consumer applications. This is the real home local area network capability to tie your PC, all of its peripherals and all of your consumer electronics together. MARTIN REYNOLDS: So this is the solution of connect being the phone to the PC and getting my phone numbers in my PC into my phone. CRAIG BARRETT: It's part of the solution. A bigger part of that solution is we exited 1999 and 2000 where the press was still beating up the competition between the handheld devices and the desktop devices. Was it going to be the mobile Internet or the wired Internet? Were phones going to take over all the applications of PCs? It was kind of a dumb debate, in reality. No one types long Word documents on their phone that I know of. You do? (Laughter.) MARTIN REYNOLDS: There's always one. CRAIG BARRETT: I know. But you want the utility of the handheld devices, whether they're smart phones, whether they're handheld PCs, PocketPCs, whatever they are, Palm devices. The real importance is that device be able to be compatible and interface with your database on your desktop or your laptop PC and be able to get scalable content off the Internet. So the debate about competition between here and here is finally turning around to these two should collaborate, they should cooperate, they should extend the capability. And that's where the industry is going now, everywhere from having the interfaces talk to one another to being able to write applications with performance primitives, so if you write it for the desktop, you can immediately compile it for the handheld and vice versa. MARTIN REYNOLDS: The idea being that all these things will, eventually, when we plug them in, work together. CRAIG BARRETT: The customers will demand that they work together. MARTIN REYNOLDS: Funny that, yes. JACKIE FENN: Now, getting that capability into the devices is one thing. The next stage is obviously getting it onto the chip level and adding more and more capability to the base chip itself, so all the chips are communicating. What do you see changing when the industry gets to that stage? CRAIG BARRETT: Well, I don't think a whole lot will change. Our industry has some very simple fundamentals behind it, and they're easily described by two words: innovation and integration. You go back 30 years and what we've been doing with Moore's Law, which gives you a bigger transistor budget each year, is you innovate new functionality in peripheral chips and then you integrate it back into the core. And every year there's new capability, whether it's memory capability, graphics capability, you name it. As your transistor budget goes up, you integrate it back into the core. So you start off with wireless capability. I mean, if you have an 802.11 laptop today, you've got a PCMCIA card that you shove in the side, and that gives you connectivity. Well, the next thing to do is to integrate that innovation, that add-in card into the core processor/chipset combination so it's built into the device. The software, the drivers, the capability are built into the device, auto configuring. So I buy this thing off-the-shelf. I go home, and the first time I find myself in an 802.11 hot spot, the thing auto configures. And that's value to the consumer, but it's the capability that our industry brings, which is that let's take that continued innovation and integrate it into the core capability on the motherboard. MARTIN REYNOLDS: So when do you think this means pretty much every PC will ship with wireless? CRAIG BARRETT: Well, we're introducing with our products this architectural innovation in the first quarter of next year from a laptop standpoint. And clearly, the desktop is going to follow shortly behind that. MARTIN REYNOLDS: It certainly takes 12 to 18 months, so in 2004 we'll all have wireless all the time, right? CRAIG BARRETT: Well, I think you will see the build out of these public hot spots. You're seeing it in the U.S. already. Many airports, restaurants, hotels are getting it. It's going to be a competitive advantage for them to have it. If you go out of the U.S., because sometimes we get too U.S.-centric, you see in Japan this is happening. Another interesting part of this is the whole 3G 802.11 debate which is happening in Europe with the telcos there that spent well over $100 billion for that bandwidth, and it's not really great bandwidth. It's 1 to 2 meg with the 3G spectrum. Now they find themselves, they're going to be competing at the local area network level with 10 to 50 meg competition. But again, it's not competition. It's really cooperation between the two. You know, most networks are a combination of a wide area network and a local area network. 3G happens to be the wide area network. 802.11 happens to be the local area network. So it's not just U.S., but around the world this is starting to happen. MARTIN REYNOLDS: All right. So in just a few years we could see this. I remember when phone lines first started appearing in hotel rooms. Days when we traveled with screwdrivers. Remember that? And what? Within two or three years it seemed like every hotel had a jack. CRAIG BARRETT: Well, the most important thing IT departments used to do was give you this little bag of tools to take the faceplates off and wire into the telephone circuitry of hotels. (Laughter.) CRAIG BARRETT: It was really tough for the folks who were colorblind, though. Remember that? (Laughter.) MARTIN REYNOLDS: Let's turn to talk about another of one of Intel's investments, the Itanium technology. How do you see that, the state of that now as compared to where you wanted it to be today? CRAIG BARRETT: Well, you'd always like your technology to move faster. We introduced the first member of that family a little over a year ago. We've introduced Itanium 2 processor just a few months ago. I think it got very, very strong technical reviews from a performance standpoint compared to any of the competition. We've been working with the entire ecosystem. The one thing you recognize in this business very quickly is having the world's fastest microprocessor is interesting, but it's not particularly useful to anyone unless it comes equipped in a system with an operating system with the application stack which are ported and tuned to it. So for the last year we've been working on the latter three aspects. We had the processor. We worked with the computer OEMs to provide the systems. We worked with the software OEMs, and whether that's Microsoft, whether that's HP, whether that's Linux or whoever that might be with, to get the OS really ported and tuned to the architecture. And then the application stack that sits on it. So we're looking for that family to ramp up rapidly through the second half of this year into '03. We have the next three generations of processors in design to follow on the Itanium 2, so every two months you're going to see another leap forward from technology, performance, capability, price/performance to the end user. MARTIN REYNOLDS: So your plan is just to keep investing in Itanium, moving it up in the market until it really takes off? CRAIG BARRETT: I'm not sure about moving it up in the market. Our plan is to keep investing to make sure that the solution stacks are available on it. We think the price/performance capability of the technology is without question. But you have to have the solution stacks. And we're gratified, you've heard Carly, who was out here a bit ago, and I think was signaling HP's full support for it. Within the last week or two you've seen Dell finally say, "Gee, we weren't so excited about the first generation, but Itanium 2 looks cool. We're going to come out with the standard two or four way processors with it. You saw IBM commit first quarter of next year, they'll have a two way processor, that basic building block for scalable systems on it. Now, you look around the world, there's really only one computer company that's not fully embraced that architecture. He'll be out here in 15 minutes. (Laughter.) CRAIG BARRETT: Hopefully, someone will ask Scott when he comes out, you probably won't have to ask him. He has so many one -liners about Itanium that he'll just spit them out. (Laughter.) CRAIG BARRETT: But it's interesting to spit out one-liners. We're more interested in bringing the technology into the marketplace in a sophisticated fashion, with all the computer OEM support in the world, multiple operating systems lined up behind it, and then hundreds and hundreds and hundreds and thousands of applications ported and tuned. We think we're making great progress in that area. And you pick your industry: financial industry, manufacturing industry, the petrochemical industry. You pick it, we're working with people in all of those industries to move it forward. MARTIN REYNOLDS: So the key to Itanium at the moment is you've got key deployments in strategic industries. CRAIG BARRETT: The key to any computer architecture is making solutions with the customers find useful. And if you can provide them increased capability, increased price/performance advantage, increased competition for the product line, they can buy from a variety of vendors, that's what customers want. So we'll keep investing on our part, but we have to work with the rest of the industry to make sure the whole ecosystem is there. JACKIE FENN: Looking beyond Itanium, we had some questions of what happens ten years out. Is quantum computing going to kick in? It's the perennial question, how you maintain that growth and at what point you have to shift to fundamentally different approaches to computing. Can you tell us a little about the research and how you're addressing that? CRAIG BARRETT: Sure. Everyone acknowledges that some day the basic CMOS transistor will be replaced, much as the vacuum tube was replaced, you know, 40 years ago. And you have to look at the fundamentals of the technology. And we scale CMOS transistors. We're at 130 nanometers, going to 90 nanometers, going to 60, going to 40, going to 25, going to 15, going to 10 over the next five or six generations. That gets you to about July 26th or 27th of 2017. (Laughter.) CRAIG BARRETT: At that point in time, this becomes a pertinent question. (Laughter and applause.) JACKIE FENN: We like to take long range views. CRAIG BARRETT: It's good to take long range planning. (Laughter.) CRAIG BARRETT: There are two important things you need to recognize, though. You know, the basic information gets stored on an electron spin on a single atom. If you trap the atom or move that atom, you can move information like a CMOS switch can move information. But there are other alternatives. The alternatives are nanotubes have got a lot of press recently. Very thin organic films. All sorts of possible new switches. That's all the transistor is, is a switch. That's all the vacuum tube was, was a switch. So we look at all of those new technologies as they come along. JACKIE FENN: You're looking at it peripherally because of the time involved? CRAIG BARRETT: We look at them in conjunction with our long term research aspects, and mostly we look at them in conjunction with where long term research is done. So major research universities around the world. The other important thing to recognize is almost regardless of what that new switch looks like, you're going to build these new circuits the same way that we build them today. They're going to have to have very small dimensions. They're going to have to be built in clean rooms of the sort we have. So we may fundamentally change the nature of the switch, the fundamental compute element, but the layers of metallization, you know, the size of the wafers you put out, none of that is going to change dramatically. You'll have those same manufacturing techniques. You may change the base material a bit, you may change the fundamentals of the physics a bit, but the construction process is probably not going to be terribly different. You're still going to have to print incredibly small dimensions. So we won't use visible light or UV. We'll use extended UV or soft X-rays, whatever you call it. And you'll build them in the same way we build them today. You know, I think I've got one here. Just happen to have this prop in case we needed it. A 12 inch wafer built with basically single gate CMOS transistor state. JACKIE FENN: Tell us where that fits. CRAIG BARRETT: That's just a bunch of test patterns on there. It's not worth much. But the fundamental construction, manufacturing technology isn't going to change a lot, whether we're building CMOS transistors or quantum dots. So the technology may change. It will be an interesting discussion topic. The issue will be to recognize that the technology is changing, and then be well positioned to take advantage of it. You know, that is the key to presumably this conference and the key to technology, the key to everything we do. You need to see what technology is available and position yourself to take advantage of it. There's always more technology alternatives than you can take advantage of, but you have to carefully pick what makes sense for you, what you can do in the future and how you can use that for the success of your company. JACKIE FENN: You, as Intel, you're very strong at driving the technology forward from the kind of bottoms up perspective. Do you try to get a vision of, yeah, what people do with all that power, how things might be reshaped based on the technology advances you're making? CRAIG BARRETT: Well, to a certain extent, you have to have a vision of the future and you have to be able to do new things and you have to, I think, be a role model of how you want the world to turn out. To the greatest extent, if you move up the food chain back away from individual companies even to countries or to nations, they have a fundamental issue of how they want their country to turn out, just as we have fundamental issues of how we want our business to turn out. At the national level the things you have to worry about are education, how you get enough educated people to drive this following toward. By the way, for those of you who are U.S. citizens, if you haven't noticed we don't do a particularly good job at that aspect in the U.S. in terms of graduating the numbers of IT professionals we need. You have to put the infrastructure in place, and that is you have to choose the computer, Internet, broadband, wired, wireless infrastructure you want for your citizens or your company to be successful. And then you have to role model the behavior that you want. And one of the big roles for governments is to role model the behavior they want, e-Government initiatives. One of the issues for companies like Intel is to role model the use of the technology and the behavior that you want. That's why we try to use the Internet to sell all of our products. It happens to be the most efficient, most productive, most customer friendly way to do it. Ditto on buying material, ditto on interfacing with your employees, ditto on interfacing with your partners or fellow travelers. So we try to apply our vision in practical reality to achieve what we want. But that's really the role of everybody here, to do that. JACKIE FENN: You mention that the vision and the international angle. I know you just come here from a trip to many countries in Europe and the Middle East. Do you see different visions there in terms of how they are promoting IT and education. CRAIG BARRETT: Yeah, I visit about 30 countries a year around the world and try to talk to government leaders, business leaders, academic leaders. You get the same message around the world. Natural resource, agrarian-based economies are out, knowledge-based economies are in. Knowledge-based economies are determined by intellectual capacity; that is, the ability to educate your work force, and the infrastructure you have in place. The cities or countries I visited last week were Dubai, Jordan, Israel, Munich, Moscow, Paris, Rome, and Seville. That's not bad for one week. But you see exactly the same issue everywhere. Everyone recognizes this. Everyone's in competition. So it's not just that members of your audience who work for specific companies recognize the competition within their company sector, but countries are in this competitive thrash as well. And they're in the competitive thrash of what do I do to improve my education system, what do I do to improve my infrastructure, and what do I do to role model how I want the technology used. And that's key wherever you go. MARTIN REYNOLDS: Do you see any countries that are doing better than others? Specifically better. CRAIG BARRETT: From an education perspective, it is very clear that the countries like China and India graduate far more engineers and especially IT professionals than the United States does. The fact is if you look at detail in the United States, the percentage of our graduate students are foreign nationals each year tends to creep up, and those people increasingly go home to their countries rather than to stay here because the infrastructure is being built up in their countries to allow those people to have a comparable living standard working at home as opposed to having to immigrate here. So I see the education sector of the U.S. is almost at the bottom of the barrel of the industrialized world. Results from the international tests on 12th graders show the U.S. in the bottom 5 percent of the top 25 or so industrialized countries in the world. We do pretty well at the university level, but we just don't graduate enough people. We still have the best university system in the world. Just about every country in the world has an e-Government initiative. Just about every country is worried about their infrastructure, whether it's wired, broadband, whatever you're talking about. So, if I were to give the U.S. a grade in this area and where we would sit in the next 20 years, I'd be pretty pessimistic. I think there's immense competition. The work force is going to shift increasingly out of the U.S. just because the education dynamics push it out of the U.S. And if your standard of living is directly proportional to the average education of your work force, that means the standard of living shifts out of the U.S., or the rest of the world catches up where the U.S. comes down, but either way, it's a pretty pessimistic viewpoint. I'm talking to two Brits here, so forgive me. But the UK is not much better than the U.S. (Laughter.) MARTIN REYNOLDS: Well, we have opinions on that, right? JACKIE FENN: Just to wrap up, we covered a lot of ground here in terms of education, technology futures, Intel's internal processes and many of the investments you've made. Perhaps you could just wrap up by giving some bottom line advice based on your unique position and the things you see, what should our IT audience be primarily focused on? What should their goals be when they get back to work next Monday? CRAIG BARRETT: I think the main challenge of any of us in business is to make a judicious selection of the salad of technology that's out there. Not every bit of technology is useful to every company. But IT executives have to work with line-of-business executives to figure out what bit of technology, what portion of technology brings an advantage to me in my marketplace against the competition that I face. So the challenge is not to immediately adopt all new technology, but it's the careful analysis of the plateful of exciting technology that's out there, what's useful for your company. What really gives you a return on your investment? And that's not an easy issue. It's why all of us face the complexity today. But it's just so much wonderful technology. You can't use it all, but how do you select what is the best technology for your company? What gives you the best return? It's clear to me we will continue to provide new technology, so we're not going to make your life any easier going forward in this area, but that is the beauty of this industry. It just continues to move forward at an incredible pace. You don't have to take every bit of it, but you can take selective pieces of it and use them to your advantage. And that's the challenge that we face. MARTIN REYNOLDS: And of course we're always here to help. CRAIG BARRETT: I thought you'd get that commercial in at the end, Martin. Yes, you're always here to help. MARTIN REYNOLDS: Great. Thank you very much. Thank you, Jackie. CRAIG BARRETT: Pleasure. (Applause) * Other names and brands may be claimed as the property of others. |
|