Keynote Transcript

Intel Developer Forum, Spring 2002

Craig Barrett
San Francisco, Calif., USA
February 25, 2002

CRAIG BARRETT: Thank you, Patrick. Those of you who were here last year will remember his introduction was a little bit different then, but he decided he couldn't tell the difference between a male or female horse, so he gave up on those sort of introductions.


CRAIG BARRETT: I have a very simple topic I want to talk to you about today, and it's really based on the following: Why the world on the heels of the worst recession this industry has seen really in its lifetime should anybody be positive and excited about what we have going forward? And I'm excited, perhaps more excited than I've ever been in my 28 years at Intel about the opportunities in the future, not just for Intel but the industry as a whole and what we can do. The message I want to give you today has to do with technology and that now more than ever is the time to really break away from a technology standpoint, bring the end user some really exciting technology.

We have a simple saying at Intel that the only way out of a recession is basically to bring out new products, new technology, new capability, and make the end user excited about what you have to offer. And that's exactly where we are today.

We have to innovate our way out of this recession with new technology, new products. If there was ever a time where we needed to move faster down the technology curve, this is it.

So I'm excited. I'm super excited about the opportunity in computing, communications as they converge, and as this technology permeates the world.

The opportunities we have are really limitless. And that's what I want to talk to you about today. It's really my perspective on what's happening with the economy, where we are, what are some of the technology areas that we really need to make significant advances. I want to continue to point out its advances, which the end user can appreciate here, the people who buy our technology. What are the areas that we really need to focus on to grow our industry, and then we'll finish up with a few trends and transitions.

Transitions in our industry are important because transitions are where market share is gained or lost. What you really need to do is focus on bringing new technology forward with each transition or each break point in the industry. So we'll talk about a few of the areas that are exciting to our industry as a whole, and especially exciting to the consumer base.

If you look at the technology cycles for the last 28 or so years, and actually this curve is consistent with the length of time I've been at Intel, there have been basically nine complete cycles, ups and downs. You can see how they're listed here in terms of oil shocks or PC bubbles or gulf wars, etcetera.

The unique thing about the one labeled number nine is, in fact, that it is the deepest trough that the semiconductor or microelectronics industry has ever faced. Obviously it's a combination of a number of things, the recession in the U.S. after ten years after pretty strong growth, the shock, the overinvestment in the telecommunications marketplace. All of those things gave rise to about a 30 percent, 35 percent reduction in the semiconductor revenue in 2001 over the year 2000. Just an unprecedented decrease in reduction in a heavy capital, R&D intensive industry.

Now, why in the world in the face of such a decline could you be positive about the future? The only reason you could be positive about the future is what our industry has in front of it is a worldwide build out of the Internet, the convergence of communications and computing, and the fact that the best is still ahead of us. That's what led us over last year and forecasted this year to spend $20 billion plus in capital equipment expenditures and R&D.

We recognize that the only way you get out of the red zone, that little red dot labeled number nine and get it back up on the growth side of the curve is new products, new technology, new capability, new excitement for the end user. You have to invest to do that.

That's a message we gave last year. That's a message we can give even more emphatically this year. Collectively, we need to invest and bring new technology into the market.

So if you look at the industry tone, I think the only way to describe it is kind of mass confusion. Almost on a daily basis the stock market goes up or down depending on one company's earnings. I visited one of our design centers in Mexico last week, so I was flying down to spend the day there. I was reading the business section of the Arizona Republic. It proudly proclaimed that the recession was over. I think the stock market went down.

No one knows exactly what's happening. All of the so-called experts who proclaim they know exactly when this recession is going to end, none of them pronounced when it was going to begin, so they've all established their credentials of not knowing what the hell is going on, so don't bother to listen to them.

What you really need to listen to is the message that regardless of what happens, the only way to get out of a recession is with new products. The unique thing in our industry is technology continues to move forward in good times and bad. Old technology doesn't sell. You need new capability to get the consumer interested in what you're manufacturing, what you're doing. So you need to continue to invest.

If you look at, from a macro-scale, and let's forget the day-to-day economics and just look at massive transitions in the world's economy, driven by technology changes, and if you just think back for a while, you can think of some of the massive changes that have taken place with the industrial revolution, with the steel era, massive production of steel, automobiles, electrification, railroads, radio. You pick your technology, whatever it is, they all seem to follow the same sort of behavior over a period of time. Technology gets introduced, it's exciting, it starts to get proliferated in the market. As it gets proliferated in the market, there's exuberance from an investment standpoint over this technology that's going to change the way everyone is going to behave, change everyone's lifestyle. You might even call it irrational exuberance.

Following that period of irrational exuberance there's a period of turbulence. That's when investments that were made tend to consolidate, stock prices go down, people gather together and say, well, gee, this technology was supposed to change the way we all behave. How come all the business models that were put forward don't work?

If you go back in the steel era, industrial revolution, railroad era, back to even canal building in Europe in the 1700s, you find exactly the same sort of scenario. Rapid excitement about the new technology, doesn't meet all of the expectations immediately, people over invest, there's a crash in the market, and then the basic foundation of the technology moves forward and, in fact, the sustained growth or the golden era of that technology happens well beyond this point of turbulence.

That's what I think we're seeing in the communications and computing sector today. The Internet was in fact the technology that was going to change the universe. All the dot-com companies were going to change the universe, even all the products were going to change the universe.

The beauty of capitalism is people are allowed to do those things, invest and waste their money, and then the build out occurs.

I think what we're looking at in front of us is this period of sustained growth. The Internet is it. The Internet is the basis for communications, for information access, for commerce, for entertainment going forward. It has just started its build out around the world.

That's why I'm confident about the future. I'm confident that this technology will continue to grow, will continue to expand, will continue to change the way we work, live, and play, and that we collectively have the opportunity to influence that.

Now, to make that happen, you need new technology: real products that do real things for real people. And it's this concept of bringing new products out that I want to continue to push.

Because you collectively, working together, provide the capability to bring exciting new technology in the market. There's not an Intel company represented here that can do it all by themselves. But what we have to do is collectively create great new products, bring them to the market.

Last year we brought over 4,000 products into the market, 4,000 new products into the market. That's wonderful, but they're basically not very useful to the end user unless we combine those products with your products, and we bring great technology into the market.

To give an example of that, last year we talked about the concept of the extended PC, as the PC being the center of your digital universe and surrounding it with other digital devices, audio devices, video devices, communication devices, imaging devices, consumer electronic devices. What I want to do is give you a short demonstration of some of the progress we've made in the last year in terms of bringing great new capability into the market that the end user can appreciate. I want to give you that example with just, very simply, what you can do with a digital camera, some cool software, and a high-performance PC. Using those three things to create what I call some real PC excitement.

So what I want to do is we're going to run a short video and then give you a demonstration of the real content behind that video.

(Video playing.)

CRAIG BARRETT: All right, David, some pretty cool stuff there, but what are you going to do to match that?

DAVID SIDD: We're actually going to show you how we're going to create all this content, but every 3D effect you saw there was created using RealViz* technology, and in the movie industry they're very interested in this technology. The special effects is interested in this technology. Special effects houses like Sony Image Works that did Charlie's Angels and the company that did Shrek are my favorites, but I heard your favorite is the Matrix.

CRAIG BARRETT: Yeah, I saw it about six times. I love the dodging bullet routine.

DAVID: I think I've seen that. But Escape, the company that's working on the effects for Matrix 2 and 3, they actually purchased the software and I'm really looking forward to what they're going to come up with.

CRAIG BARRETT: I'll look forward to it as well, but Matrix 2 is not today. Today is Intel Developer Forum, and I talked to you two guys a week ago and I said do something really cool for today, right?

DAVID: Yes, exactly. And we brought Temis here, and what we wanted to do is have the digital still camera that I'm sure all of you have, and the power of high end personal computer and software from RealViz, which just so happens to be an Intel Capital portfolio company. Shameless plug.


CRAIG BARRETT: What's your stock price doing, huh?

DAVID: Okay. So we're actually going to have Temis here to show us what he did using a Pentium(r) 4 processor tech demo system running at 3 GHz, air cooled. So Temis, using the movie, you would need location, you need props, actors, director, big budget, a lot of time, and if you were to do this in a traditional 3D animation, you would still need the time and the money. But using RealViz technology, you can create content quicker and easier, and they use something called image-based content creation. What that is is innovation that allows you to capture the beauty and the richness of the real world by using your digital still camera, and create a virtual world that you have total control over time and space in.

CRAIG BARRETT: Let's see some of this.

DAVID: Okay. Well, Temis, let's start off with the location. How do you do that with your software?

TEMIS: To do that I went to Embarcadero here in San Francisco and took a series of photos with my digital still camera and I brought these into stitcher where I blended them together into a seamless environment, and this gives us our virtual location.

DAVID: Beautiful. So you have your location. Now you need a couple props, right?

TEMIS: No problem. I took a series of images with my digital still camera of this public train and I brought them into the image modeler. With image modeler I created a 3D model of the train. As you can see, the model of the train retains all the detail and richness of the surfaces of the real train.

DAVID: Incredible. So you got your location, you got your prop. So you have your virtual set.

TEMIS: No problem. That's easy. I just load up scene weaver, and I can import the location I have in the stitcher that I have here, and I can also import the prop I created in the 3D image modeler. The great thing about scene weaver is it can define and set up how I want these props and location to interact with each other.

DAVID: Great. So now he has all that done, export this to multiple file formats. What he's showing here is Shock Wave* 3D, and this is a very interactive photo realistic environment that you can post on your Web site. Very nice. But Temis, we actually asked you to create a movie for us.

CRAIG BARRETT: Some excitement.

DAVID: Some excitement.

CRAIG BARRETT: Some action.

DAVID: You came down a couple days ago and spent about ten minutes with Craig here, took a couple of pictures of him. He's probably curious of what you did with that.

TEMIS: That's because the real beauty of the whole process is that with just a digital still camera you can take a few pictures of a person and get a very life like 3D model of the person, and once you have that, you can have it interact with your props or locations in any way that you like. To show you what you can do with this technology, I created a little movie, and let's take a look.

DAVID: Got to hit the start. There it is.

(Playing movie.)

DAVID: All right, all right.


DAVID: So what you saw there was completely all computer generated. No video footage was used. All they used was a digital still camera, a high-end powerful personal computer and software from RealViz. This is what they can use to create content for games, Web content, and of course blockbuster movies.

CRAIG BARRETT: Okay, guys. That's great. That's a perfect example of what I'm talking about, technology break away, excitement to the end user. Nice job.


CRAIG BARRETT: My part of that was, in fact, two minutes of getting photographed with little things in my face so they could match it up somehow. But it only took about two minutes of my time and then creating the content with the software. But that is the sort of excitement that the PC can bring to the end user in the future. That's the sort of thing we collectively need to make happen.

If you look around the world today, at times I get a slightly different impression when I read the U.S. press. I get to travel to around 30 countries around the world to talk to different leaders. And what happens in all of those conversations is the same thing. Every country recognizes that its future depends on its ability to create value, to participate in a knowledge industry.

I was in Mexico last Friday and talked with government leaders there. It's not that they want to continue to drive their economy with a "Made in Mexico" logo. They want a "Created in Mexico" logo. They want to move up the value chain. If you go to Taiwan, you see the same thing happening. Every country wants to participate. In fact, in the next economy, it is going to be a knowledge-based economy driven by value creation, technology is the key to that. Even though we have a heavy investment here in the United States and I.T. infrastructure, it's always good to realize the United States represents something less than 4 percent of the world's population and that the rest of the world is trying to catch up. The necessary ingredient for the world to catch up is I.T. infrastructure, which means it's a huge opportunity for us, really an industry infrastructure as well as a consumer infrastructure.

If you look a little bit at the numbers here in terms of the infrastructure that's in place driven by the services as a percent of GDP as some of the significant economies, whether it's USA, Germany, India, or China. You can see what the infrastructure in many countries looks like today as compared to that in the United States or Germany. The opportunity for growth around the world is huge for all of us.

Every government recognizes this is what they need for economic security. This is what they need for economic development. This is what they need for economic competitiveness. In fact, I would contend that most of the countries recognize this, even more so than the United States from the standpoint that we have the infrastructure in place, we take it for granted, it's assumed here.

This is what is going to continue to drive the world from an investment standpoint and capability standpoint. It's this worldwide opportunity. It's an opportunity for all of us as well.

Now, if technology is the growth engine for the economies around the world, then that means it's incumbent upon all of us to reduce that capability and to spread it around the world, participate around the world. Many of us are seeing this in our business models today.

Intel's business has gradually moved out of the United States. Some 70 percent or so of our revenue comes from non-U.S. based revenue at this time, and I expect to see that continue to grow.

From a breakaway standpoint or technology breakaway standpoint, what is important for all of us? I want to emphasize this concept of technology breakaway, because this is really what is necessary for us to thrive going forward. It's really bringing new capability into the market. Not just trying to sell what we have today, but worrying about what are the requirements next year.

What are the capabilities that the consumers want next year? What capability will business want next year? What do we have to offer? What investments are we making?

Most of this is going to be around the Internet. My contention is the Internet is it. It is the convergence of both communications and computing. There's one Internet. You can address it wired, wireless, hand-held devices, cell phones, PCs, laptops, a wide variety of devices.

There's one Internet, though. It is the vehicle for information access. It is the vehicle for commerce. It is the vehicle for communications and entertainment. The beauty of those four items is that there can be interactive, coupled together. They all benefit from the same thing. They benefit from bandwidth. They benefit from compute power. They benefit from applications, rich user interface. They benefit from all the things we do.

To continue to drive the Internet, to continue to drive this change in the structure of basic fabric of the way we live, work and play, what we need to do is to continue to bring technology innovation, new capability, new performance, new uses, new users.

This is what we're all about. You know, we have maybe 400 million people attached to the Internet today. We should have 1 billion in a couple of years and it will continue to grow. This is what it's all about for us.

Basically bringing excitement, new technology, new capability to these areas of information access, communications, commerce and entertainment. This is what is going to drive our collective business going forward.

Now, if you look at the foundations of all of this, what is going to drive it, you know, semiconductor technology. That's where we come from. If you want to continue to drive performance, then you need to be confident that you can continue to drive this sort of thing that Gordon Moore forecasted in 1965. He initially forecasted the number of transistors would double in circuitry every 18 - 24 months. That got changed sometime ago that the processing power would double every 18 months. This is, in fact, what has been driving our side of the industry and helping that drive our cooperation with many of you in the audience over the last couple of decades. This is going to continue at least for another 15 years or so.

You know, we will make more and more complicated devices. We'll have the ability to put billions of transistors running at terahertz speed in a single integrated circuit. The overall device will run in the 30 GHz range. It will have transistors as small as 10 nanometers. Not many of you know how small that is. It's small, 1/100 of a micron.

Gate oxides today are measured in this thing (showing object) which has three atom widths of silicon dioxide in it. We can see this progression.

The basic CMOS transistors die and it's filled with nanotubes or something like that, the basic technology to build these switches. CMOS transistor is a switch, if you're going to replace it with another switch, the basic technology to build those new switches is the same you make today.

You'll still have to wire up hundreds of millions of billions of them. You'll have to make hundreds of connections in the outside world. That's the sort of thing we do today. That's the sort of thing we're on track to continue.

Now, I want to show you a very simple example of what we're doing at Intel in this space in terms of 300 millimeter wafers. We talked about this a little last year when I was here. I invited a photographic team to go up to our Oregon facility and show you what goes on. The first part of this video is not too exciting. They just got enamored putting their bunny suits on and the way they had to shield the particles from the interior of the cleanroom, and I think they also wanted to take photographs of each other for posterity.

We're going to cut this, and I want to take you inside the cleanroom now and show you what it looks like in a modern billion dollar plus 300 millimeter silicon wafer factory.

(Video playing):

D1C is one of the cleanest places on earth, about 1,000 times cleaner than a hospital operating room. At only one particle per cubic foot of air, critical for creating electronic circuitry on the scale of human DNA. The automated handle system virtually eliminates the need to handle wafers start to finish. The 300 millimeter manufacturing process involves more than 300 steps creates 24 billion transistors, with some circuitry only 100 of atomic layers thick and takes about a month to complete.

The microprocessors created in the 300 millimeter process incorporate 0.13-micron technology versus the 0.18 micron technology of its 200 millimeter predecessor. A 2X leap consistent with Moore's Law.

The area of the 300 millimeter wafer is over twice that of a 200 millimeter wafer. Yielding 2.4 times the number of microprocessors.

Combined with the effects of Moore's Law, D1C's output advantage is more than four times that of a 200 millimeter 0.18-micron facility. These technological advancements enable quicker time to market to meet customer demand and reduce per chip cost by 30 percent.

Utilizing Intel's Copy Exactly methods, D1C's manufacturing process for the 300 millimeter wafer will be duplicated in fabs around the world beginning with Fab 11X in New Mexico.

D1C, the world's first 300 millimeter fab, another technology breakaway from Intel.

CRAIG BARRETT: These 12-inch wafers are kind of cool, but I'm looking forward to the next generation that are going to be 18 inch in diameter and we're only at a small size pizza today and I want to get to the large size pizza going forward. But the critical thing that was mentioned in that video was the reason for doing this. This is a batch process. What you do is get a substantial cost reduction in any batch process by processing more area, so the cost per unit area goes down 30 or 40 percent if you can double the area of the wafer you're using.

So you go from eight inch to 12 inch, you get over a doubling. You go from 12 to 18 inch, you get the same sort of thing.

Technology's there. We know how to do this. We can continue to do this going forward.

That's our form of technology break away. It's basic silicon technology. What do we do collectively together, and what are some of the areas for sustained growth we need to work on going forward?

Security is one. All you read about in the newspapers and trade magazines today are the issues of security. How can you have trusted transactions? How can you secure your information? How can you ensure that only you and intended parties have access to the information? What can you do about broadband? How can we get the pitifully slow adoption of broadband in the United States to speed up? What can we do to make that more interesting, more exciting? What regulatory issues are there? And then what technologies, both wired and wireless technologies, come forward in the broadband case?

What do we do about Web services? Lots of talk, different companies. Microsoft's got some very exciting aspects here. Other companies as well.

How will these pull together? What will the applications be? What does the infrastructure need to look like? What do you need to do to get hardware acceleration of Web services, get the end user more benefit, more capability?

How do you move wire to wireless computing? How do you have your compute mode go anywhere, anytime at high broadband capability?

I think these are four of the areas where we can collectively work together to bring the end user greater capability, greater excitement. These are the sort of problems that we need to collectively solve. The technical solutions are basically there. In many instances, it's rolling out the technology or coupling different pieces of the technology together. Certainly from a broadband standpoint, the basic technology is there. It's really the infrastructure and the roll-out of that infrastructure that is important.

What's our ultimate goal out of all of this? The ultimate goal is basically to bring technology anytime, anyplace, anywhere for any individual. And that is the combination of the PC, whether it's a desktop or laptop, the PDA, the handheld computer, the cell phone, whatever other access device you may be talking about. How do we get all of these to work together in a seamless fashion and give the end user more reason to want to use our technology?

For a couple of years we were engrossed in arguments about whether it was going to be the PC or the cell phone, which of those two was going to dominate the universe. We didn't stop to think that it really wasn't a matter of which one was going to dominate the universe; it was really a matter of how do they work together to give the end user more benefit?

When the end user has more reason to want to access the Internet or communicate device to device, then we all benefit, we all win and the technology moves forward.

So this has to be our ultimate goal. You know, the technology for everybody, anytime, anywhere.

Trends. I think most of you are aware of these trends. One trend is, in fact, computer segmentation. You'll hear a lot about that from the presenters following me. It used to be one size fits all in the compute world. Now the characteristics of the handheld device are very different from the characteristics of a laptop computer, very different from the characteristics of a desktop computer, different from the characteristics of a workstation.

Computer segmentation means you have to design devices and characteristics segmented to that particular application, whether it's power consumption, performance, user interface, whatever it might be.

Open standards. This industry has been built on open standards, built on open interfaces that the whole industry can innovate around. That is how we marshal the entire force of our industry to innovate together. That has to be a prerequisite for what we have going forward.

If you look at the communications industry, there's been a rude awakening in the last 12 months. Actually, I think the communications industry in the last 12 months has experienced about 20 years of the computer industry learning in the space of just 12 months. That is that modularity, standard building blocks, standard interfaces is going to be as important to the telecommunications industry as it has been to the computer industry. This movement of the computer industry to standards and modularity standard building blocks is going to be a trend which there is no turning back from and a trend which will provide an opportunity for many of us in the audience here to provide those building blocks.

These are the key trends we have going forward.

Volume economics, worldwide distribution, end user value, segmented to the absolute need and requirements of that particular device, and being able to take modular building blocks and put them together to build something wonderful.

Computing and communications are merging. No question about it. The important thing for all of us to recognize is not to fight that merging, but to make it happen more seamlessly.

How does my 3G or GPS cell phone, how does my PocketPC interface with my laptop, my desktop, to the one Internet, to the one database that I have? How do I get scalable information over whatever bandwidth it may be, wired or wireless, to the device of interest with the different types of displays those devices have.

So this whole concept of computers and communication merging, we've been talking about it for years, it's finally getting there as we roll out higher bandwidth into the mobile side, we roll out high bandwidth in the wired side, and we work off one Internet with scalable content. This has to be key to all of our product definition.

If you look at the sort of things that we do inside and the sort of things you're going to hear about this week, they really fall into three buckets: clients, networking, and servers.

You'll hear about two forms of clients this week, really, desktop or laptop, clients or handheld clients. Wired or wireless clients. You'll hear about networking. Networking is really the plumbing that holds the Internet together. Increasingly going modular, increasingly going to standard building blocks, increasingly going optical.

What about the optical electronic interface? You'll hear about some of that today. What about standard building blocks for network processors? You'll hear about some of those issues tomorrow.

In the server space, those big things that hold all the information in the world, direct all the traffic on the Internet, take the typical data center, divide it into three categories of Web servers, application servers, and database servers. They get peculiar requirements for each area we're talking about. High performance 32 bit, high performance 64 bit capability. You'll hear about our plans for 32 bit servers, you'll hear about our plans for the Itanium processor family. Mike Fister is going to follow me this morning. He'll talk about the five next-generation Itanium processor devices that are under design today as we move farther into that space.

It's really this Internet computing which this slide is a model of, from the clients to the network and to the server, encompassing all aspects of the Internet, these are the areas where we collectively need to work together. And you can do some wonderful things as you build this up, as you build this capability up. What I want to do is just show you a very simple demo in terms of what you can do with high bandwidth and a high performance PC.

We have lots of discussion about why we need more than a gigahertz? What do I need more than a megabit of bandwidth for? Let's so what you can do with high performance digital TV, high bandwidth, and a big microprocessor.

Jeff, why don't you come out and do a little demo for the audience.

JEFF: Thanks, Craig. What I'm going to show you is what you can do when you have a robust broadband architecture in place and enough processing power on your PC. In this setup here, we began with our carrier class Xeon server. We then served it under 100 megabit, which is processed on a Pentium 4 Hyper-Threading technology enabled small concept PC.

CRAIG BARRETT: I'm still waiting for 100 megabit to get to my home.

JEFF: It will be there.

CRAIG BARRETT: I have about 12 now.

JEFF: This piece of technology here is running at about 3 GHz.

CRAIG BARRETT: Another 3 gig demo. What can we do with this?

JEFF: With the help of software from accessDTV we can software decode not just one but two high definition streams simultaneously. What you see here are two high definition streams with picture in picture mode without the use of any high definition acceleration hardware. Each of these streams is coming down at a data rate of about 20 megabits per second and it's able to handle it.

CRAIG BARRETT: You can do something better than that, can't you?

JEFF: Since this takes about one gigahertz to run one and we're running about three gigahertz here, let's embed all three streams into a 3D environment. So what we have here are three high def. streams coming in from our server at a data rate of about 60 megabits per second. Software decoded on our client and embedded into this 3D environment.

CRAIG BARRETT: Is that me on my snow board?

JEFF: That's you. So now what you see here is just a glimpse of what the power of the Pentium 4 processor is enabling for application, Web, and game developers. Now, what we have here is a broadband architecture that allows you to embed the highest quality video content into your apps to create a richer, more exciting user experience. And all this is delivered using Intel architecture building blocks.

CRAIG BARRETT: Great. Thanks, Jeff.

JEFF: Thanks, Craig.


CRAIG BARRETT: There was a lot of discussion about broadband policies around the world, and basically I think except for the United States, all the other G7 countries have broadband policies or are starting out very strongly stating what to do in broadband.

A number of organizations in the United States, TetNet, which is located here, have put out statements of guidelines for what the U.S. policy should be on broadband, and it basically says trying, within the next decade, to get 100 megabit connectivity to 100 billion homes in the United States. That's the sort of rich content you can get with that kind of broadband capability. If you have the question about why do you need a three gig processor, basically, just remember what Jeff said. It maxed out at about one gig with one stream with that processor. You can take the state of the art processor almost to its knees with a couple of streams coming in.

That's some of the excitement, some of the capability going forward.

Let me really conclude with the importance of targeting transitions with innovation. Really, transitions are where you gain and lose market share. Transitions are where you have the opportunity to move from a recession to a positive, vibrant economy and benefit from that positive vibrant economy.

If you look at some of the transitions that have taken place recently, Intel goes from a Pentium III processor family to a Pentium 4 processor family. Microsoft almost coupled that transition going from their current operating system to Windows XP*. Those are the sort of exciting transitions and capability that drive the market.

What we have to do collectively is to drive our product transitions to user-perceived benefits around the world, whether it's wired or wireless, anywhere, anytime, capability across the board, exciting new technology. That's what we're all about. That's what we ought to be working together on: Bringing these new exciting technologies in the market. That's how to get from the bottom of that chart showing the growth of the semiconductor industry or the computer industry or the communication industry, to take that upward and get on the positive growth element.

Working together, we can do that. But it needs a continued investment in technology, the continued establishment of standards, modules, modular capability, building blocks that we can all build off of going forward.

As I said at the start of my presentation, I have more confidence today than anytime in the last 28 years about our capability to move forward, to change the world and have the Internet really be it. By "it," I mean the standard for communications and computing, the standard that's going to change the way we all work, live and play. Let's go and do that together.

Thank you.

* Other names and brands may be claimed as the property of others.