[Due to the varying sound quality and subject matter of tapes, the information in this transcript may contain inaccuracies. Downloadable Word document and Netshow files on-demand are located at the bottom of this page]

MR. GATES: Thank you. I understand you're having a great conference, and I'm really impressed with the attendance we've gotten. I thought I'd spend a little bit of time talking about our overall strategy, and then give everyone a chance to ask whatever questions there are.

It's a very exciting time in the software industry. The growth in the business, the innovation, is really fantastic. And, as was mentioned, there's a move away from people developing their own software, and using packaged solutions that share information across the Internet. And one of the things that drives all of this is the rapid advances in the hardware capabilities. Microsoft was founded based on the vision that this doubling in performance every two years would lead to machines that would be both inexpensive, and would be powerful enough to deal with the world's most complex problems.

And we see the pace of those improvements being even faster today than ever before. In the next couple of years, the clock speeds will continue to go up. We'll get up over 1,000 megahertz, gigahertz Pentium chips from Intel. We'll also see the introduction of the next generation processor, which is the 64-bit Merced. We've been working very closely with Intel, NT actually runs on their simulator today, and we have a lot of special tools that are going to make it easy for people to move their software over to the 64-bit environment, whether that's on Merced or also on the Alpha chip as well.

We're very enthused about the new systems designs that are symmetric multiprocessing using many CPUs. Breakthroughs in bus speeds are now allowing us to move up from the two and four processor systems to eight, 16 and even 32. And so overall systems performance will be much higher.

We're also really excited about the benefits we're providing through clustering. By clustering systems, we can get the fault tolerance that people like Tandem have done in the past on non-PC systems down into the PC space. And clustering is also a very important technique for getting up to the very, very high-end performance.

Microsoft has a lot of products, but there are themes that go across everything we're doing. The first theme that you heard a fair bit about from Bob Muglia is digital nervous system. This is a very important idea, because it's speaking to businesspeople about the value they can get out of their information systems if they embrace it fully, if they move away from paper-based information flows, if they move towards empowering their workers with the very best tools.

For the consumer, our theme is Web lifestyle. This the idea that, over time, people will expect to be able to get information to do banking, to learn new things, through the Internet and companies have to get going to make sure they're on the Internet and reaching out to existing and new customers in that way. So, those are the two things that are really changing the way people think about using the technology.

Of course, our major products are going to stay the same. There's been really very little change in this list for a long time. Windows is our biggest product, Office very close to that, BackOffice is our fastest growing product and a place where there's a lot of innovation. The tools, Visual Studio, and then finally the pioneering work we're doing out on the Internet to understand what kind of tools are needed there, and to really try and develop that market as rapidly as we can.

Across all of these products, there are a number of initiatives. Internet support has been something you've seen us really put a lot of priority on in the last couple of years. Whether it's getting the browser integrated into the shell, getting the Web server integrated into every copy of Windows NT, getting our Office products to work very well with HTML, I'd say a very high percentage of our work has been enhancing Internet standards so that people can do easy information exchange through those capabilities.

The payoff on that has been really quite fantastic. The Internet is not a fad. It's something that every year just simply gets more and more important. And I have no doubt that, even though in the United States this is further along, the benefits that people are seeing here are really global in nature. And so, no country will be outside the changes that are coming through the Internet.

A second goal is scalability. We have a lot of things working in our favor here, not just the software improvements, but all those hardware changes I talked about. And today a single server can do pretty unbelievable things, even very, very large businesses can run a lot of their software against the single server and for demands that go beyond that connecting them together is getting to be easier and easier.

The transition of writing code on the mainframe down onto PC technology requires us to make it simple for people to do that in stages, and that's why interoperability is so important. And we've come a long way in our products coexisting with UNIX machines, with mainframes, and any of the legacy devices that are out there.

Finally, the one that I think is probably the most important, and really there's a lot of invention that we need to do in simplifying the product, making it so that people don't see error messages as much, or so they don't have to learn nearly as many commands. That is crucial to not only grow the market, but allow people to get at the power that these systems have, and so that's really something that our research work will help us out with quite a bit.

Digital nervous system is a term we've used for just over a year now. I was meeting with some CEOs last year and trying to explain to them how some companies are really going to step back and get more out of their PCs and network connections than other companies, and that's when I came up with this term. In fact, I'm working on a book that goes through some of the business benefits and some great examples of companies that have moved toward this very rapidly.

Things like electronic mail are just the building blocks that go into this. You have to have employees comfortable with using the computer, you've got to have the Internet connections, not only inside the company but to the partners. So, by and large, the hardware base required here is nothing more than what you would have even if you were just using your PCs in a very limited way for productivity software. The main thing you need to add to that PC base to get 100 percent of the value out of it is standard applications packages, and some thinking about what information really counts for your business.

And there are some metrics that we talk about. You know, can you get all the customer information in one place? Can you see everybody who works with that customer very quickly? Can you use the Internet as a new way of getting out and sharing information with your customers, and making it easy for them to do business with you? Do you have an information system that allows bad news to travel fast? Companies really need to know what's going on out in the market, and most particularly they need to know problems that are taking place.

Here at Microsoft, people sometimes send electronic mail saying, we've won a certain account and, you know, isn't this great. And I don't mind that kind of mail. But, in fact, when you get that kind of mail, there's really nothing you're going to change. You know, it's good news, and it means let's just keep doing what we're doing.

On the other hand, in most companies, you don't get too much mail where people are saying, hey, we lost this account. But that's what you really need to know about, because it might change what you're doing. You need to know about competitive activity. You need to know about customer feedback that says things should be better. And so, a good way of using the digital system is to make sure that that bad news not only gets into the system, but it gets to the people who need to know about it.

This vision is really about empowering workers, giving them all the information about what's going on so they can do a lot more than they've done in the past. Also, I think there's a basic philosophy here that by empowering these workers you'll make their jobs far more interesting, and they'll be able to work at a higher level than they would have without all that information just a few clicks away. Simple things like being able to go back and see what happened in the past, do easy comparisons. On paper-based systems, that's almost impossible.

I don't think any company today fulfills the whole vision of digital nervous system. Microsoft is trying to make itself a showcase, and there's a lot of great examples that are out on our Web site or in this new book I'm working on that touches on all of this. But there is a huge opportunity for the industry to show people how to get value out of these systems, at the same time that we improve these systems and lower their cost of ownership.

Now, our success has really been based on partnerships from the very beginning. We decided back in 1975 not to be like other computer companies where they not only did their software, but they did the systems, and the chips, and everything all in one company. By specializing in systems software, we were able to have a focus there, and to work with all the different hardware companies that believed in the PC. It meant that we were very dependent on our relationship with Intel to build the chips, with systems builders, particularly like Compaq and HP, to make the systems, and with all the value-added service companies that would have to be there to help people take out building blocks, our software, and put them into full solutions.

Now, a lot of our success comes from the partnerships with software developers. When you think about an operating system, it doesn't do much by itself. And really, the only reason people pick a particular operating system is because applications run there. And, applications have to be very broad. We need literally tens of thousands of applications to cover all the different requirements all over the world. And pushing those standards, getting the information out, working with developers, has been something that has been important to us from the very beginning.

What you're seeing at this conference, and some of the things Chris just talked about, is that there's a renewed emphasis on our work with developers. And in particular, a huge emphasis on developers who build applications that focus on particular business requirements. We have a lot of great horizontal applications, but when it comes to the breadth of what we need, we still have a lot of new companies that we want to see come along, and companies that are taking their expertise from previous platforms, not PC platforms, and now bringing them down into the PC space with this great new performance.

To work with partners, we have a lot of different programs, certifications that are important, or support offerings. The key themes in all of this is, we know what we're good at, and it's very specific. And so, we want to work with people who are first-class in their area as well. And we really measure ourselves by how does it come together with customers.

A few examples that have been great over the last couple of years include SAP and Baan. SAP a few years ago was primarily supporting very high-end systems, particularly UNIX systems. And by working together to make NT a good platform, bring up the performance, now the majority of their sales are on Windows NT, and of those a substantial portion are on SQL server which is something that they just started about a year ago. With Baan, now they're making Windows NT, their total focus for new implementation, and they will be doing some great stuff on top of SQL server because they have some very special features that take advantage of unique extensions that come out in SQL Version 7.

And so Windows NT has incredible momentum for not only outselling all UNIX put together, but we're growing at a really quite amazing rate. Over 80 percent in the last year, and that kind of incredible growth will go out into the future.

SQL 7, you'll find this is a big focus for us. We're very excited about the new release that's coming out late this year. The beta 3 that is the final beta is in the hands of over 50,000 people. If you're interested in that, it's something we're making available to all developers who are interested. We've got some very good migration capabilities, and some great performance. The performance, of course, is the combination of software improvements and hardware improvements going together. This just sort of gives you a sense of what's happened over the last 30 months.

Back in the start of 1996, our performance on this industry standard benchmark was about 25,000, and the cost per transaction was about $250. Eighteen months later, with a four processor system, the performance had gone up very dramatically and, in fact, the price was down under $50. The year after that, just recently, we took SQL 7 and the latest hardware that's available now, the Xeon-based servers that many of the people are coming out with, in particular a Compaq system, and got a benchmark result of over 18,000 transactions. And so both on a relative basis and absolute basis, that kind of curve is pretty incredible.

That kind of improvement will continue out into the future. And so, the notion that people think of the PC as something that's limited, or specialized, that really is going away. In fact, a lot of our focus now is much more on the fault tolerance and extremely high reliability. That's an area where we've got some clear metrics. We're really tracking everything that goes on there to make sure that people think of these NT systems as not only as reliable as the previous systems, but more reliable, because with the very high volume we have, we can afford to invest in the testing and development at a level that no other software platform could possibly receive. We're spending today over a billion dollars on the Windows platform. And that's been going up at a rate of over 20 percent a year, because the volume allows us really to be very, very aggressive at building new features.

Charles Stevens talked to you about what we call ADCU, our applications development customer unit. This is part of the new emphasis we have on working with developers, and partly it reflects the notion that we need to reach out to developers, particularly in Europe, in a way that we haven't before. We need to make sure that all the expertise that we have had in the United States we've got great people who can meet face to face, and help out all of you, wherever it is that you're based. And so we've grown this group quite substantially. We've gotten into the verticals and tried to understand who the key players are, and worked closely with them. And we're very pleased with how this is going. And of course, an even like this is part of the plan that we have there.

I mentioned that our development costs are increasing very rapidly, because of the opportunity we see. Our R&D spending is getting pretty close to 20 percent of sales. In fact, the only thing that holds us back from getting to that level is the challenge of hiring in so many people, and maintaining the very high standards that we have here.

But, whether it's the testing, the development, all of those things, we're being quite aggressive, and even when you compare us with other technology companies, this emphasis on R&D is very, very strong. Even someone like Oracle, whose business model is very, very similar to ours, you can see quite a difference there. Our total R&D budget, across all the products, would be a little over $3 billion this year, which is of course another big increase.

Now, part of the reason for this is that we really see no limit to what we can do in making the platform a lot better. Things like using the network connection to automatically update, automatically fix problems that go on, to let somebody across the network help you out if that's necessary, making it so that you don't have to think about your mail program stores information one way, the file system another way, Internet pages are stored another way. Bringing those concepts together, so you just learn one command, even if underneath the system is handling them in a special way.

We've got to get to a natural interface, an interface where a computer can understand your speech, a computer, if you call it up over the phone, can talk to you in language that you not only understand, but it's enjoyable to interact on a speech basis. We'll also have the computer using these low-cost digital cameras, to recognize who's using the computer, what kind of gestures are they making, what are they looking at, what kind of reaction do they have to what's going on. And really make the computer far more efficient to work with, not making the keyboard the only way that you get information into the machine.

Now, based on this, we're going to make it so you don't have to think about where the information is. If you update your schedule on one computer, it should be immediately available to you on all your other machines, without having to do anything that's a special command like you have to do today. So if you go somewhere, where there's a PC, as soon as you authenticate who you are, all of your customization, all of your information will be there, and you can do you work, even if you haven't brought your own machine with you.

The storage system inside the PC will be advancing very rapidly. The differences between the file system and the database will really go away. All of the information will have the rich type of look up, indexing and query that we normally associate just with the database-type environment. And so I'm sure you understand from all of that, that the system will continue to grow in its power, and we're very dependent on the chip manufacturers to continue to make memory inexpensive, and to deliver the performance, because those natural interfaces, whether it's speech or vision, do require performance levels well beyond what current chips are able to deliver.

So I'm really going to close on a very positive note here. We see incredible opportunities. I think the software industry will be the industry that is creating more jobs, providing more innovation, providing more great success stories than any other industry. This is the best industry to be in. The breadth of great software products that's needed by all the customers out there is quite incredible. And even categories that you would have thought of as being stable, or well defined, are changing quite a bit, because the Internet redefines how people are going to expect information flow to take place, not only inside a company, but outside the company as well.

We want to make it clear how serious we are about Windows. How serious we are about working with all of you. And we're very anxious to get your feedback on what our priorities should be in getting you the information, on helping you out in the marketplace, so together we can seize the huge growth that's out there.

So thank you very much for being here, and it's great talking to you.

MR. CHRIS ATKINSON:I'd like to throw the floor open to questions now. If you can put your hands up, if you have a question, and if you put your hands up, we'll basically get people to come over with cards. And if you have a question, keep your hand raised, and we can then cue the questions up from the cards.

So I believe we have our first question, number three, at the back.

Can you give us your name and company just so we know who you are, and then please go ahead and ask your question. Thank you.

QUESTION: Good morning. My name is  -- (inaudible) -- it is my perception that Microsoft Exchange has been evolving and  -- (inaudible) -- in the marketplace, from a few email applications, at the time it was launched, to a true and useful, really mission critical, messaging and  -- (inaudible) -- platform, which it is today. The success of LANS, WANS and the Internet are driving the business needs for really powerful messaging platforms, like Exchange here. So we think that unified messaging and CTI's functionality is the natural following step -- (inaudible) -- in the marketplace, and solutions like the one we developed in  -- (inaudible) -- which are in line with this trend. I mean, it uses Exchange as its messaging platform of choice for unified messaging.

So the questions are, if you agree with this vision, how do you anticipate the unified messaging market, needs and evolution, and how Microsoft is planning to bring Exchange to cover these needs. And the second one is how does Microsoft plan to leverage -- (inaudible) -- to capture these developing marketplaces, like unified messaging and CTI, and what new opportunities will be there for companies like mine, that are totally invested in driving Exchange deployment as the unified messaging platform of choice?

MR. GATES:There's a number of good questions in that. We are incredibly serious about Exchange. In fact, if you look at the time our people in the field spend making sure customers pick Exchange, it is how they spend more time than on any other activity, because Exchange has the directory that is the Windows NT directory, in NT version 5. Also, Exchange, the front end that connects to Exchange Outlook, is an integrated part of Office. And so we're evolving so that the boundaries between the Office applications really pulls Outlook in.

So anybody who is using Windows or Office really should pick Exchange, and of course there is intense competition there with Notes, where over the last six months we've done quite well. We've been winning more design wins than they have, by an increasing amount. But, it's still something that's very crucial to us.

When you talk about unified messaging, I think you're talking about the desire to include voice mail, and those capabilities, in with a single mailbox. So your faxes, your voice mail, and your electronic mail all come to one place, and you can manage them in a simple way. I totally agree that over time everybody is going to expect that. And it is something where we think there is a big opportunity. We have enhanced our APIs for telephony, they're called TAPI. TAPI version 3 is a very big advance, which is a platform for anybody who is doing applications that want to interface with the phone. Any sort of customer management, sales management applications would want to take advantage of what's going on there in TAPI.

There is a question when you get into spaces like small business of whether over time we can make the NT integrated package we call Small Business Server rich enough so that you actually get TBX functionality in your NT box. That's not something that will happen in the near future, but actually the idea that your data network and your voice network become one is something that could save money and give you additional functionality. And certainly, partners like Cisco are pushing pretty aggressively to have the kind of performance and low cost that could move to that single network.

In terms of our particular interface with you and attacking this unified messaging opportunity, you know, we really should sit down and talk about that. This is an area where as Exchange evolves there will be changes in the way that you tie into the product. Hopefully we can make that work better. And so I think we need to have some very direct dialogue about the specifics of what you're doing. And certainly, we thank you for your focus on Exchange. And we'll follow up on that.

MR. ATKINSON:We'll take a question from microphone number two.

QUESTION: Yes, my name is  -- (inaudible) -- the mobile market is growing extremely fast. Probably the fastest growing operating system is the operating system in the mobile phone, and it's been there -- (inaudible) -- from Motorola, and from Nokia, that they're having less in the company volume -- (inaudible) -- and this new company called Symbion (sp) which was created around that.

I just wanted to -- this is something which is, of course, quite a challenge over Microsoft, in terms of volume. Will Microsoft withdraw from that market, in terms of Windows CE, or will you put more effort to compete or try to get into this market? I mean, the figures for mobile phones is 250 million units are going to be sold in the Year 2000. And it's almost the same size as the PC industry today.

MR. GATES:You're absolutely right that mobile telephony will continue to grow. The percentage of those phones that actually have rich data capability will start out fairly low, I think. But, then as the costs come down, and some of the key functionality is broadly accepted, I think, certainly in the next 10 years, that will become standard capability. Microsoft is very serious about this area. We really have two big thrusts. One is that when people put in the network infrastructure that they use Windows NT for the systems that create that mobile backbone. We've got some very good partnerships there, and we're pushing ahead very aggressively on that.

The second is the part you focused on, which is the software that runs in the phone itself. And there, our strategy is to take Windows CE, and its ability to understand directory entries, its ability to understand schedule entries, ability to understand office documents, and tailor that for the mobile phone environment.

And despite the fact that that's an incredibly competitive environment, with probably over 10 different operating systems all trying to get into that space, we're actually increasing our efforts pretty dramatically. And, in fact, I'd say if anything, the creation of Symbion will cause us to reach out to some new partners, and do more, more aggressively than we might have done otherwise. We really believe in mobile data, mobile data is going to be a huge application, whether it's for the PC or other form factors.

And, in fact, the Palm PC form factor, where you've got handwriting recognition, that and the phone, you'll see some very innovative form factors that some of these new partners are working on, and that will include GSM support, as well.

MR. ATKINSON:Okay. Thank you.

Can we have Microphone number four now, please?

QUESTION:  -- (inaudible) -- and we're just launching in France and in Germany, a solution, collective solution based on SQL server and -- (inaudible) -- my question to you is we are in an area that -- (inaudible) -- when I talked to the partners here, and with all the ISVs, we all had the same kind of suggestion to improve the way in which relate to Microsoft. What we need to understand is not just the feature coming in new products, but so that we can commonly add more value, and focus our investments where you are not going to add value in the next version. One of the key factors is to understand what will be your plans on the different products, for the next releases, beyond the ones that you are planning at the moment. I think this kind of dialogue is very, very important to leverage all of the investment and the response capacity that we have, within the network of ISVs.

MR. GATES:I think that's absolutely right. You take NT 5, I think we did share very early on in the process what our plans were for NT 5, what we wanted to build in. The areas that we beefed up, as the product has been under development. We beefed up the intellimirror, because of the cost of ownership requirements, and we beefed up the applications richness, the building the transaction server in, and moving up to this COM-plus level, where the services are not something you have to pay extra for, but come with the operating system. And that's partly based on developer feedback, and wanting to make it easier to build distributed applications.

In terms of where we go after NT 5, the big themes you'll see are going to be storage unification, making it so that the way you store pages, and files, and databases, and mail, all of that really, from a developer's point of view, and the end user point of view, all of that looks like one rich store. This is a vision we've had for a long time. And we now are very enthused that we'll be able to do that in the major release that comes after NT 5, which will be certainly 2 to 3 years after that. Another big area for us will be starting to pull down in support for speech recognition, natural language understanding, and that's a huge emphasis for us, because we don't think every application developer ought to work there.

In terms of particular areas, you know, we should sit down and have a dialogue. I think a model example that we need to recreate in other areas, is what we did in NT 5 with storage. You know, there's a lot of rich things that go on in storage, where people do tape libraries, and hierarchical management. We enhanced the NT 5 storage, but we were very clear what Microsoft was going to do, and we did much better multi-volume management, and put that into NT 5. But, then we worked with the outside companies to put in the hooks that they could do their value added pieces on top of that. And so even though there was an imperative for us to do more, we I think struck the right balance, and communicated very well about what's going on and our plans there. And in each area we need to have that dialogue.

And the  person you work with in ADCU should be able to get you in touch with the right technical group at Microsoft to get that dialogue going. It's certainly to our benefit to make sure you're not taking your resources and duplicating things and that, in fact, you're building things that make NT a stronger platform, and benefiting from the volume. Although we're doing that well in some areas, I think it's something we need to do more of.

QUESTION: Good afternoon. I'm Roy Thornton -- (inaudible) -- and my question was about knowledge management. What do you mean by knowledge management, and where does it fit in Microsoft's priorities?

MR. GATES:Well, the term knowledge management is a very popular term now that is used differently by different people. I can give you sort of the Microsoft view of the term, which is, if you have information in one part of your company at some point in time, how easy is it for another part of the company at a different point in time to easily find that information if it applies to what they're doing? And so, today, when the information is just filed in a hierarchical file system, even with the kind of indexing support that we provide, it's not good enough to make sure that people can find the previous experiences and learning that went on inside the company.

And so, for us, knowledge management means making it very easy to navigate to the previous documents or messages or records that might be relevant. And we're extending Exchange and Office to work into those scenarios. You know, internally at Microsoft, we have a lot of knowledge, and it's a good-sized company. And so we're trying to make ourselves a real test site for all the new things we're doing in this area.

It's obviously a term that encompasses things that Microsoft will do, and then a lot of things that will be specialized packages built on top of the platform.

MR. ATKINSON:A question at Microphone Number 3.

QUESTION: (Inaudible) -- there have been rumors recently about delays in the release of NT Version 5, can you promise us a delivery date of the product?

MR. GATES:Well, I can give you an update on where we are with NT 5. We've been very clear from the beginning that our top priority with NT 5 is the quality of the product. And so, in fact, given that as a priority, we don't know what the exact date is, because once we get our beta 2 out, we'll just listen to those beta sites, of which there will be over 200,000. And based on their saying, hey, this thing is really ready to go, that's when it will ship.

We did have a delay in beta 2. We were hoping to get beta 2 out by the middle of '98, and now it looks like it will be in August when we'll actually be able to ship beta 2. So we had a little bit of a slip there based on wanting to get a lot of the functionality into final form, and get a very stable beta 2. We expect that the timeframe after beta 2 will be typical of what we've seen in the past, and what that means is that we think that it's extremely likely that the operating system will be out in the first half of '99, but we cannot give a specific date prediction. What we will do is, as time goes on, we'll explain exactly what we're hearing from customers, and give you the prognosis based on that.

MR. ATKINSON:We'll take a question from this microphone here, please.

QUESTION: Hello. My name is Afredo Christoforitia (sp). I'm working for Rockwell Information. And part of our business is the software business and integrating data from the plant form to the -- (inaudible) -- system. But our traditional business is programmable logical controller, PLC systems.

Now we are seeing a great impact of Windows in the manufacturing industry, and I would like to ask you what is your vision, especially now that you are releasing -- (inaudible) -- so what is your vision, what is the impact of CE on the manufacturing industry?

MR. GATES:Windows CE is aimed at a very broad set of applications, and so what we're doing is making Windows CE available to lots and lots of companies for them to build it into different devices. Some of those are consumer-type devices, like building it into the car, which will take several years to be high volume, but eventually could be a very exciting thing. Some are these hand-held or palm-sized PCs. And then another whole area is the industrial area that you were bringing up.

We have had a very positive reception from companies in the industrial automation area to take Windows CE. Part of that is that in going from what was a fixed logic device in the past to a general purpose microprocessor, the cost of the hardware has come down so much that even having a general platform like Windows CE is now very economical. And so, we're working in our vertical area to define standards for how these different machines can talk to each other, and what APIs are in Windows CE for people to build manufacturing applications on top of that.

It will really be through the key partners that this rolls out. I'd say that the demand there has surprised us, it's faster than we expected. You did mention the real time. Right today, Windows CE has very high performance and can meet most requirements for manufacturing environments, but it's actually the next version, which is the 3.0 version, that has the hard real-time guarantees. And that's not out until some time in the first half of '99. But we have already gotten a lot of input on what those extensions should look like, and we're hard at work on those. So, I think it will replace the dedicated logic devices as people build new manufacturing lines, simply because the flexibility is there, and that the costs are very low. And we're not going to be the experts in that area, and we certainly look forward to working with you and see where you think this can fit in.

MR. ATKINSON:Thank you. Can we take questions from a speaker at Microphone Number 3 now, please.

QUESTION: My name is Gavin Kilful (sp), and because of the toughness of the question, I'm going to leave out the company that I'm with. The question revolves around, as an exceptionally busy executive, I know that you don't get much time for relaxation. And having seen you on certain great advertisements I wanted to ask the following question, and that is, using your new Big Berthas and your Calloway golf clubs, are you finding mastering golf more of a challenge than running Microsoft?

MR. GATES:Well, I find golf very relaxing. In fact, this Saturday, I'll be out with some friends. You know, really for me it's a way to get away from work and get outside, and mostly it depends on, you know, what the group is. That's what makes it a lot of fun. You know, I don't take it nearly as seriously as I do my job at Microsoft. If I did, I'm sure I'd be very, very frustrated since it's not a game in which you can get great consistent results, which is certainly the standard I apply to the company. We expect to basically, you know, win championships on a regular basis, and so far we've done very well at that. But, its a lot of fun, and once you get going it's almost kind of addictive.

MR. ATKINSON:At Microphone Number 3 again.

QUESTION: Good morning. I'm Mike Sparian (sp) from -- (inaudible) -- I think we all agree technology is evolving very fast today, and maybe Microsoft technology is even evolving at breakneck speed. But the theory is that real customer demand is a little slower. And we independent software vendors are the bridge, which is needed to bridge the gap between leading edge technologies and real customer demand.

I believe Microsoft is wonderful in supporting technically the software vendors. Do you have a plan in the future to support us even to -- (inaudible) -- the way customer demand may evolve, so that the gap can be smaller, if you have -- (inaudible) --

MR. GATES:Well, I think we are in a position to get out some key messages, and that can make a big difference in terms of whether people think the additional software investments make sense or not. Take, for example, the impact of the Internet. We're sending out a very strong message that you probably ought to get modern applications that will let you take advantage of the Internet. And in every industry we want to take showcase examples where people have bought the right package, have organized their business that way, and they're getting ahead of the competition as a result of that, either better product decisions, better customer service, better partnerships, finding new customers. There's a lot that can be done for people who are using technology in a better way.

And so, I'd say our marketing support for you is at several different levels. There's the broad message of Microsoft about technology. There's the specific marketing that we can do within the vertical area that you're in. And then there's the question of whether there's specific marketing things that we can do together. Part of the reason we're having the application developer customer unit have people out in every country is so that we can sit down and talk through what those opportunities are. For us, there's a big emphasis now on applications that are moving up to take advantage, not only of NT, but also the BackOffice, particularly exchange and SQL. And because those are very competitive areas for us, that's where we'll put a lot of priority on the joint marketing that would go out there.

In terms of the overall atmosphere of customers' willingness to buy software, I do think one of the key issues is lowering the complexity of keeping the software up-to-date. This is an area where we haven't done enough in the past. NT 5 is a big step forward in this area, being able to distribute updates through the Internet, there's a big update there. But so far people have viewed it that if you're going to get a new piece of software from us, and perhaps from others as well, that it's going to be very disruptive. And by lowering that perceived cost, and by making it clear what the business imperative is, which is the whole digital nervous system theme, the idea of working through the Internet, I think we can keep these markets very, very healthy. And we'd love to have a direct dialogue about these different types of marketing and how they might help you.

MR. ATKINSON:We're going to take a question from Microphone Number 2 now, please.

QUESTION: Good afternoon. (inaudible) And we operate largely in the financial and banking sector. I would like to echo some of the questions put earlier on about NT. And I'd like to focus specifically in one aspect. Is there a strategy for the development of NT in the future as an alternative operating system for large mainframes, such as, say MVS?

MR. GATES:Well, the answer is absolutely yes. We think that the transaction requirements of Internet servers will actually be much greater than anything that people have asked mainframes to do the past, or what they'll be able to do. And we also think that the price performance, and development tools and integration people would get by using Windows NT will make it far more preferable. Now, this is not the kind of thing that -- where people make an abrupt transition. But, there already are examples of companies, like Microsoft or Bridgestone, or some medium-sized banks, that are running large enterprises, that historically would have used a mainframe, totally using Windows NT technology to do that.

For most companies, the question is when they write new applications are they going to write them on the mainframe, or are they going to write them on NT. And part of the reason we've done such strong gateways with things like the transaction server gateway up to CICS is so that even fi they want to keep their data on the mainframe for a much longer period, they can move their application development down onto the NT platform, and then at a later date decide when they want to move their master database along with that. So the interoperability strategy is really a way of decoupling those two decisions. And you know, when we look at performance, when we look at reliability, we are absolutely -- have very specific plans to go beyond the performance of very high-end UNIX, very high-end mainframes, as well as match the kind of reliability that those systems have been able to deliver.

For us, it's an imperative because of new applications. The total number of mainframes out there isn't that large, but if we're really going to get people comfortable with using NT and knowing that it can scale better than anything else, that's going to have a huge impact. And, you know, it's very natural given the investments we're making, that we'll be able to do that. I'm always surprised that there is skepticism about this, because it's just a matter of engineering, and it is a very key priority for us and we're putting a lot into it. So I think the people who are skeptical will be surprised over the next couple of years, with the clustering and the focus we've got here, that we will be able to deliver on that.

Now, even once we deliver it, people won't immediately move to take advantage of it. It will take, you know, five years, six years, before you see a very high percentage of customers having new applications that make them look at switching to a new platform.

MR. ATKINSON:Bill, we only have a very short time left now. I have a question submitted by email, which I think certainly will speak for a number of people in this room. If you imagine yourself to be the CEO of an independent software vendor, developing line-of-business applications, what would your top technical priorities be in the coming years ahead?

MR. GATES:Well, I think the very top priority would be thinking about how all of this connectivity, the Internet connectivity, is changing the way people want to use applications. After all, applications in the past were very oriented towards running in batch, on a computer, and you have these paper reports that come out. They weren't oriented towards triggering electronic mail, or feeding their data into a spreadsheet, so somebody can look at it in a pivot table. They weren't oriented towards letting, say, your accountant connect in through the Internet and look through things, as opposed to having to come on site and actually go through the paperwork. And so there's absolutely a revolution taking place, in how you design the applications. So the first thing would be the issues around the Internet, and how the NT platform makes it possible to get out and do those things.

The second area is the overall modularity of the applications. The trick here is to be able to write the code so that unique customer requirements don't require you to go in and make a lot of changes. And Microsoft is providing COM as a way of separating the pieces of your application, even making it easy for one piece to run on one machine, another piece to run on another machine. And so understanding how I take my existing code base, and get it into a more modular form, that would be a big priority, because eventually I'd like customers to be able to do customization, without my having to go back in, and have them be able to accept new updates. So that whole modularity approach is a very big thing.

I don't think that it's necessary to rewrite applications in a new language in order to get either the Internet support, or the modularity. The Microsoft strategy of supporting all languages, C, Visual Basic, Java, and making sure they all can scale up and take advantage of the new platform services, that means you don't have to rewrite all your code. You can simply structure it in new modular ways, add the Internet functionality, and you're set.

MR. ATKINSON:On behalf of the audience, we'd just like to thank you very much for being able to join us this afternoon. Thank you.

(Applause)