MR. GATES: Good afternoon. It's great to be here because it's certainly an exciting time in the computer business. The shifting ways that people are thinking about the computer is creating immense opportunity. It's only this year that you can see for sure that businesses of all types are using the computer for internal and external communications as the top competitive issue within almost every business.

One of the dynamics that has truly driven this industry has been the growth of PC volume, the improvements in the chips, the wide variety of software, the innovation, and manufacturers of various sizes. It's really led to a pretty incredible device. And if you think of the PC five years ago-which almost no one would use today, -- that pace of innovation is faster than ever.

People talk a lot in this business about open versus proprietary. The dream of openness isquite a simple dream. It's the customers as they go to make future purchases. They don't have to think about what they have bought in the past. The only place where that's truly happened, in the entire computer industry and in its total history, has been the PC itself. People can choose from hundreds of manufacturers, the latest innovations are as outright as people want them and everything you think of, whether it's peripheral costs, software costs, anything costs less when it's associated with the PC than with devices that don't have that kind of volume.

Now, one of the biggest dynamics is that this PC success story is moving up to the server level. This has happened as the performance of the microprocessor is there and as the application developers are now moving their products, which came from main frames, UNIX devices and minicomputers onto the PC. And so that's led to lots of companies investing and making the PC platform the best server platform; the best in price, performance and flexibility. And again we are talking about servers that come from dozens of companies. Compaq makes great servers, HP makes great servers, but they are not the only ones that will use the latest technology to offer up these devices. True openness is here rather than in the space where the term is used on such a constant basis.

Microsoft's role in this has been to take Windows NT and constantly use the customer feedback loop to drive it as a more robust foundation for mission critical applications. The role of NT was quite ambitious: to take everything that was great about UNIX, the high-speed filing and the Windows compatibility, and pull that together. Now, more and more, having integrated server application is very important.

It was always a surprise to us that Novell never developed a general purpose server application platform. They stayed with very specific vertical services, file sharing, directory, and so today if you want to take an Internet or Internet application that typically involves a fair amount of code up on the server, a fair amount of scripting or data rules to be executed as a part of dynamically generating those pages, it's not possible to do that on a network server because it doesn't have the preemption and the development tools and the critical map that NT has been able to develop.

Windows NT volume has more than doubled in the last year as the distribution channels and application and all that momentum came through. NT 4.0 has been out about a month now. That was a major step forward. The metric multiprocessing is where we are tuning and tuning to get more performance. Obviously, as Pentium Pro is moving into the mainstream, we get a huge benefit there.

Two things we are working on now you will not see until next year are the clustering or scalability and the 64-bit support. We have a lot of great partners driving that. People like Tandem, Compaq, DEC, many of whom bring experience in doing this type of clustering with other types of hardware, now bring that down into the PC space.

The basic concept here is that instead of thinking of your server as being so many different services, you think of electronic mail and administering that, and you think of file and printer sharing and database access, Internet access. All of these-you have one directory, one-to benefit all of those.

All software development is going on assuming this broad very high-speed NT network. It's not everything it needs to be yet. It doesn't have the quality of service guarantees that will allow you to drop your private network altogether, but already it provides virtually universal electronic mail conductivity and browsing conductivity.

The phenomenon is one that people have sort of thought, "this is going to happen over night." In fact, there are a few things that hold it back like the speed of the connections-particularly out in the homes. We are investing as much as we can to promote ISDN and PC cable modems, but even in five years' time if you take, say, Europe and the U.S., at best 30 percent of homes will have those high-speed connections.

The key point is that like the PC itself, here we have something in critical mass where the more people who get involved, the more material there is. The more material there is, the more people will come in and get connected. So the pace of innovation is pretty fantastic. When we look at how quickly these communication infrastructures will improve, we have to keep in mind that in parallel with that, the hardware innovation continues to be quite rapid. The speed of the processor is doubling every two years while the price of the PC continues to come down. No problem there.

People are going to have the speed to let them do the rich graphics and the new input desktop; no doubt the investments being made there are starting to pay off. The real limitation of "would you rather read something off the screen or read it in print?" has been the screen - and that's getting better and better.

Today it's a dilemma for me: "which do I read? The trade magazines on the screen or in print?" because it's still a little bit easier to browse what is in print to make sure you see everything. So right now I'm reading both, which takes even more time, but I suspect that over time that will migrate to be purely electronic. DVD technology, you get to the rewrite form, DVD RAM, that's a big breakthrough, because it's the first device that's a super setup, the floppy disk, the hard disk and the CD-ROM, so we can go from having three mechanical devices in a PC to only having one there.

The form factors are going to be quite diverse. What you carry in your pocket, what you have in your car, those will be very different from what you have on the desktop. At the desktop you will want to be able to see a lot of text so we will probably continue to call that the PC, although it's changing rapidly. The other devices where they have the alternate form factors, we will give them different names and we need to make sure we have clear standards for the size of the screen and the richness of the power so that as people are authoring applications they are able to know which of these they can target.

There is no free lunch in terms of whether an application being designed for a large screen will automatically run on a small screen. You have to do work to make that go. An application that's designed for a system has a hard disk and large memory. If you are really fully exploiting that and the kinds of things it can do with animation, it will not run well on pocket-size devices. But development tools can hide some of those differences. Some material will be authored for only one of the platforms, some will go to the extra effort to hit on multiple platforms.

Intranets have kind of exploded but they are very hard things for people to measure because back when people were talking about file sharing, the goal was the same: information at your fingertips. What we have done here is simply say that the way you find files, the way you search them and put them out on the server, that there is a better way to make that information easy to get to and that's to create pages that can be browsed.

So when you go to the finance page, you see who to send email to if you are confused, you see when it was updated, you don't just see a bunch of file names. As part of that navigation you can go right down to a spreadsheet, use the spreadsheet capabilities and go back up to a web page.

Because the Internet, where you just read information, is very pervasive now, people are forgetting that most people in corporations don't just read information, they create information and view it in unique ways, so you need richer tools than just calling up HTML. HTML pages are great and they should be mixed in, but if you want to annotate a document and recompute the business plan, you need more than what today's HTML is able to provide.

That integration is the opportunity. Because customers can use today's networks and today's PCs, this is probably the best feedback they have ever gotten in data processing. It's simply taking the vision of why they got those PCs and connected them together using a little bit of additional software and not forcing their employees to learn new tools or interfaces. Then, across all the different scenarios, they have information that is easy to get to.

There is a lot going on right now with cost of ownership. This has been a topic that the industry has a hard time dealing with because it really depends on how somebody is using the PC and there are a series of best practices that make a big difference. It's easy to say if you never updated, the software cost of ownership tends to be fairly low.

That would have been true for a PC three years ago and is true today. But when people paint a scenario where you don't update the software, you have to keep in mind that there are great new things that come in that software. Take browsers alone. Will people want to use the new browsers or will they simply stick with a machine that has that browser hard coded?

Our direction is to get rid of the space that's on the PC itself so that all of the space of the PC is up on the server. If your PC breaks, you can just plug in a new one, and that state is brought down. Most PCs will have hard disks so we can cache that information very effectively, and there is a portable device used portably and, of course, we have to do a fairly complete job of bringing the information down onto that hard disk. But what we call "manageable PC" means that there is nothing unique on that local hard disk. All of it can be restored or administered directly from that server.

It's taking PC technology and taking the Intranet and using it to attack the cost of ownership issue. And it's very complimentary with all the things being done on the hardware side. So the product strategy we have is focused on use and focused on Internet integration, which we all embrace and extend, and using that integration to make sure you always have the latest software. If you are trying to print, the latest driver would come across without you even having to know about it.

Some people will say, "Printing, look at that. That's 20 percent of the cost of ownership. Let's tell people they can't print, give them a device that doesn't have that feature." That's fine for somebody who has been using the dedicated terminal, but for people who are used to authoring documents and working off-line, that's probably not going to be the solution. We will probably have to give them better software technology in order to fit into that.

The access to help is another feature. On the Intranet you can share your computer screen very easily. In fact, it can be done on the Internet as well. We have our FreeNet meeting software that let's you take the word processor and connect up your lawyer and negotiate a contract through screen-sharing, and only one side needs to have the application.

That same scenario applies to any sort of collaboration, including helpful collaboration. So whether it's access to a Microsoft support center with our vendor expertise or simply to corporate help set, that ability to see what is going on, help guide the person through, update their software, know the state of that machine. That's a very big deal. And if you take that, together with the Web-based administration, it means you can be location independent. You can go out to that remote branch and know what is going on without any difficulty.

Now, the browser has been a fast moving area. It's interesting that every six months now we have had new browsers. Eventually that's got to calm down so it's easier for people who do authoring and the software updating, but for the next year I think this pace will certainly be maintained.

Internet Explorer 3.0 just came out about six weeks ago and that's where we brought in active control. We are already talking about the next major version, which will go to beta before the end of the year, which is the Internet Explorer 4.0, that's got the active desktop. It takes the active controls and brings that to not only the desktop but all the folders you have in your computer.

So just real quickly, I would like to ask Andrew to come up and give us a glimpse of what it means to take the browser and put it at the center of the system, make it so that information you care about comes to you rather than your having to go out, browse those links to get at that information.(Product demo.)

MR. GATES: Great. Thanks. As Internet browsing becomes integral and becomes part of your mail activity so that going out and getting information and having information sent to you, you can really think of that as unified; one document format, one way of following link, going all the way down to the desktop itself, including all the folders and letting you have the richness in those documents that active controls can provide. You will see that the whole way we think of the PC will be fairly different.

Some real facts about this: the browsers are changing fairly rapidly and the browser is the largest application running on the PC. It uses more memory, more disk, than all the applications out there by a factor of two. It's also more PC intensive.

Bringing down interpretative code and bringing up images that have to be decompressed. It is what pushes the boundary. We have gone from spreadsheets being the things that really pushed the computer to the limits, and now the browsers are. And by comparison, the Office products to the application are fairly small in terms of those requirements.

We do see a need to let people get at those browser improvements. We see a need to integrate the browser with the tools where people create and interact with information, where they do more than read it. We also want to be sure that what's been good about PCs, their interactivity, is preserved. We don't want to go totally back to time-sharing. We are going to more of a balance in terms of CPU cycles on the clients and the servers. Particularly as you think of dynamic pages.

When you go to a web site, you don't want to see the same thing everyone else does. You want to see what is interesting to you. Say a Macintosh user comes to the Microsoft home page, at least half of that should be things that are tailored to their Macintosh interests and how we have done business with them. That means the server-instead of just going and getting a file and dumping it out onto the network-- will be actually looking up the customer profile and executing some rules that say for this customer, their background, what should we put into the page. That's quite a bit more CPU intensive, but part of the utilization of the web that will keep servers fairly busy.

So we have to make sure both in terms of CPU over-load and latency in the network we don't go backwards to the days of time-sharing. You see this in the most extreme fashion in productivity applications and entertainment applications. You are not going to have a game that's played where you are seeing the latency go back and forth. And in terms of the consumer markets, that's fairly important and that's a place where you won't have the high-speed connections very soon.

Portable computers are still a very key, big thing. The major opportunities here I think are quite straightforward. You have the opportunity of everybody getting connected up. When you get the PC as a communication tool, that drives penetration both in business and education, and eventually into the home. more than anything else ever has. So electronic mail is a central and richer and richer application that is integrated with the Internet.

The idea of extending out to customers then, taking information and letting orders be processed that way, letting customer service be better that way. That is very straightforward, the tools are here to make that happen. That's what's going to be driving hardware sales, communication sales and mainly application sales within these next three years.

If we look out a little bit further than that, say into a five to ten year time frame, people can say, "Boy, what about this is worthwhile? What about all that great power? What is it all going to be used for?" And the answer is very straightforward. When people look back on today they will say, "Well, this was the period where computers not only had those small screens but the computers couldn't listen, they couldn't talk, couldn't see, they couldn't learn." And those are things, those are the key barriers to ease of use. As long as you have to type in Http to do this, dot that and some other thing instead of vaguely indicating a subject you are interested in, the PC is not going to be that easy to use.

Now, what does it take to make computers do those things? It takes software. It takes significant R&D investment in software. It takes perseverance to get those things right, but the process is there and the performance that will allow those things to be done in an excellent way-that is becoming available through the progress of Moore's Law. So I see the PC getting richer and richer until it really can be a device that fulfills all of those promises. Thank you.