Steve Ballmer Speech Transcript – Microsoft Windows Server 2003 Launch

ANNOUNCER: Ladies and gentlemen, please welcome Microsoft CEO Steve Ballmer.

(Applause.)

STEVE BALLMER: Well, thank you all very, very much. It’s a real honor and a privilege to have the chance to be here with you today, and the chance to share some of our enthusiasm, perhaps most importantly some of our customers and partners enthusiasm, for Windows Server 2003.

This is an interesting time. I’m getting asked by the press particularly, is this the right time to introduce a new server product? IT budgets are down, or at least not growing. What do you really think? And I’ll tell you, I think this is absolutely the right time to be bringing incredible new innovations to the marketplace. And I think our customers really share that point of view. We’ve been excited about this product. This is one of the most significant pieces of work we’ve ever done. And certainly one of the most significant pieces of work we’ve ever done the most significant piece of work we’ve ever done — in terms of what it means for IT professionals, and for the data center.

But we weren’t sure in this climate what to expect. And as we started the process of launching the product — sharing it out with customers, putting out beta releases, working with customers with whom we’re doing joint development projects, scheduling these launch events — as excited as we were, I can still tell you we’ve been incredibly surprised by the level of interest and enthusiasm in this work.

The notion that we’ll have over 200,000 people attend one event for a product targeted for the data center server type product is incredible. The number of customers who have downloaded the product and are using it is impressive. The number of sites on the Internet that are already, from free shipment, serving up web pages from Windows Server 2003 have surprised us.

So, while the question I get asked may be, is this the right time, the reaction that we’ve gained from our customers and our business partners has been even more enthusiastic and interested and involved than the reaction we’ve seen to everything else. And particularly with a piece of work that is this important to our company, and is this significant a stride forward from where we’ve been, we think that’s a really important thing to note.

As we bring this product to market, and as I’ll talk to you about, and as our customers will talk to you about, the things that we are doing with the product, where we’re going, and what we’re changing, what we’re enhancing, I think it is important to put it in a context of what’s been going on in our industry.

If you take a look, the last five years or so have been very unusual from an IT spend perspective. IT spends out on a very, and in fact a cheaper ramp of increase in the late ’90s, both because of Year 2000, and because of the dot-com bubble. IT budgets were rising very, very quickly. People broke up their spend in budgets and plans around that. When the economy turned down, the dot-com bubble burst, and we got through Y2K, what happened was, quite frankly, the rate of increase was bigger than most businesses were still willing to build. And what happened is, projects that people really wanted to do, that were planned and budgeted, were still there to get done. But the amount of money that was available for new projects, as compared to managing and taking care of, enhancing older projects, maintaining older projects was significantly reduced. It has put our customers in the IT business in a real tight jam. It’s not that most businesses today don’t want to do new things. Sure, the dialogue is all about talk, talk, talk. But that’s in an environment where businesses and the IT people inside those businesses still want to do more.

So the challenge isn’t a challenge just of cost reduction. The challenge is a challenge of letting people do more with less. To be able to go ahead and do those new projects that are very important for business success, but still have those done in an environment where budgets are tight and where money needs to be allocated to maintaining the existing pool of applications. An important phenomenon. When people ask what will happen with IT spend in the future, I think we’ll get back on a more gradual and predictable rate of increase. But now is a particularly tough time.

Doing more with less, and as we think about Windows Server 2003, we really think we’ve got the right product for this time. I can’t tell you that when we started efforts to play here we knew we would be in this economic position. I think the theme for us ever since 1988, when we first started building server products, has always been to try to give people incredible capability with incredibly good economics in terms of total cost. And Windows Server 2003 really continues that trend.

Today, there’s three products that we’re highlighting that we bring to market. The first one is Windows Server 2003 itself, and we’re going to talk a lot about that product today, and a lot about the customers who are already using it. A new version of VisualStudio.NET, which supports Windows Server 2003, and is frankly just a better version of VisualStudio. NET. We’ve had now a year-and-a-half to refine the original release that we put in the marketplace. And that’s been advantageous. Then we’ve got a new version of our SQL Server Database, really optimized to take advantage of Itanium, and we’re going to talk about that, and some of the exciting results that customers are seeing with that new version running on the Itanium version of Windows Server 2003.

Then, as we talk about this release, one of the things I want to make very, very clear from the get-go is, this is a very significant piece of work. Some of you may have seen on the trivia questions that we were putting up there during the warm-up, that over 5,000 people working on Windows Server 2003 for a number of years. In the background of this slide, you can see this list of a number of the new additions, features, and enhancements that have gone into this product. It is not just a small, incremental release of the operating system. It is breakthrough in terms of what it means, in terms of its built-in security and reliability. It’s breakthrough in terms of the improvements we’ve made in manageability, and it is breakthrough in terms of what it will mean to software developers, and to people inside businesses who want to share information, collaborate, and communicate with one another.

I encourage all of you to actually take the time to really go through the materials out on our Web site, that our sales people share with you, to understand the depths of the product. There’s a lot in here that will help many of you with business problems that you are attacking, and certainly for those of you who have servers running today on, for example, Windows NT 4, I think that in any one of the myriad number of ways you will find a compelling theme given the depth of capability in this product to take a look at the upgrades from NT 4 to Windows Server 2003.

The theme of the release is “Doing More With Less.” Really delivering on that notion of doing more with less. Probably nothing gravitizes that statement more than what you will see if you take a little bit of a historical look at what’s happened in the Windows environment, and in the server environment generally, in terms of performance and price performance during the course of the last seven years. Seven years ago, if you took at an Intel Server running the Windows NT Server Operating System, we could deliver about 4,000 TPCC benchmarks, and we could do that at a cost of about US$110 per TPCC. Do more with less, performance is an easy way to measure it. It’s not the only thing you measure. People are interested in capability and functionality, but look at the performance. Seven years later, we have the ability now to deliver over 500,000 – 500,000 – TPCCs on a single scale-up system this is a scale-out multiple systems. It’s one single system scaling up, at a cost now of less than a cent, or about a cent, I guess I should say, of the cost of seven years ago. More with less.

And later on today, I’ll be joined up on the stage by Paul Otellini from Intel, and we’re going to talk then about some incredible work that Microsoft and Hewlett-Packard have done together with the Superdome hardware where you will see what I will call the most amazing breakthrough results that even goes beyond this in terms of scalability, performance, and price performance. I’m going to save that for Paul’s talk, but it reinforces this theme of more with less, more with less. Less cost, more output. Number one performing server operating system, and you’ll have a chance to hear about that later.

The second aspect of more with less is essentially equipping our IT professional customers to be able to accomplish and manage and take care of applications, more applications at less cost. We think Windows Server 2003 should improve the efficiency of people managing those systems by at least 30 percent versus where we fit, and by even larger numbers compared to some of the other systems that people have implemented in data centers.

We think that compared to competition, we can build that in about half the time, and with about twice the performance, given the new .NET development utilities that are built into Windows Server 2003. We think we give incredible new capabilities that will allow customers to enhance the productivity of the people inside your businesses who use information for a living, and we’re going to hear from a number of our customers, including Honeywell, on that point.

Scale and simplicity. Perhaps last but not least. This is the absolute highest quality Windows Server product ever. Each year the bar that our customers put for us in terms of their expectations on quality, on reliability, on security continue to ratchet up. And we think what you’ll see with Windows Server 2003 is really a huge step forward on a number of these important quality dimensions which are of increased importance to our customers.

As we built Windows Server 2003, one of the things that’s been very important is that we make sure that we step back and ask ourselves, how do we take a fundamentally different approach to the creation of this product, its design, its architecture, its process, in the way we interact with customers to ensure a quality breakthrough? And I think there are important aspects on three dimensions.

First, in terms of customer connection, we had over 100 customers involved as joint development partners in the creation of this product. We also have Windows Server 2003 deployed on over 10,000 production servers. There are over one million beta copies in the market. And we’ve taken a technology which has been very important to us in our Office business, a technology that we call our Watson’s Implementation System, everybody here has probably gotten one of these messages where you send an error report to Microsoft. You can say that’s a bad thing, but actually having that data allows our developers to really understand significantly what we need to work on to do the best possible job improving quality. We’ve built that same instrumentation into Windows Server 2003, and in usage during the development of the beta phase, we’ve been able to really go after the quality issues which are most significant to our customers.

From a product perspective, we made major design changes in the way we think about world-based usage of the product. The work that we’ve done on fundamental quality and reliability, we think we’ve got about an eight times improvement in reliability and downtime compared to Windows NT 4. Windows Server 2003 has software update services, which allows and helps facilitate the easy deployment of patches and fixes to Windows Server Systems. We worked out our development process, we made management and manageability of the technology a top priority. We have over twice as many developers working on the management and manageability infrastructure in this release compared to any other release. We’ve shared back all of our instrumentation data with our partners, partners doing subsystems in servers, so they could enhance the quality of their work. And, we effected fairly major architectural changes and improvements in the IIS 6.0, the Internet Information Server part of the system, enhancing its reliability, enhancing its security, and enhancing its performance. A major breakthrough on the fundamental design and architecture of that subsystem.

The other area of quality I want to talk about because it’s so in the news, and so tattooed on our brains in terms of importance, is security. We’ve gotten the message from our customers loud and clear, security is absolutely a top, job one issue with our customers. And people probably have read a little bit about the direction that Bill Gates gave our development team a year-and-a-half ago to really go forward on what we call Trustworthy Computing.

We got a number of the best scientists in our research organization to apply new thinking on the kinds of tools and technologies that we can build to help ensure code security. We’ve focused in on a number of pillars versus the code be secured by design. We have a whole scheme now that’s modeling new threat concepts against Windows Server. We stopped and we did overviews, aided by these new tools from our research group. We’ve rearchitected, as I mentioned, the Internet Information Server. We’ve changed the way the defaults in the system work, so that by default there’s much less surface area available for a task. We’ve turned off unnecessary services by default. We’ve made it so that the Internet Explorer component in the Windows server is not available to scripting, et cetera, so that it does not present itself as a place of potential attack by attackers out in the Internet.

We’ve done a lot of work in deployment, including the Software Update Service Technologies, that let’s you automatically deploy whatever patches and fixes are necessary, and we’ve done a lot of work communicating and helping explain the approach that our customers who deploy these systems need to take to ensure that they get deployed in a secure fashion. This has really been an area of focus. We think we’ve spent hundreds of millions of dollars improving the overall security of the system. People will say, does that mean there will never be another issue? And I can’t say that. I can’t say that. I can say there will be fewer issues, and there will be better technology, and we have built better processes to help respond, and to help you respond to any kinds of issues that do, in fact, come about.

Let me turn to this concept now of enabling IT efficiency. I told you, I think we can help save people about 30 percent of the cost. Where is that going to come from? It’s going to come from three primary sources. Number one is the big push amongst our customers to consolidate servers, and use servers to consolidate and take out the hardware cost, but most importantly management costs. We’ve got a lot of technology we’ve built into Windows Server 2003 to facilitate server consolidation.

Automate the management of the system. You’ll see, as some of our customers describe the things they’re already doing with Windows Server 2003 that we’ve built in new technology to help automate the management of Active Directory, the management of the servers themselves.

And, third, one of the hot topics, as I said, amongst all of our customers is security. We have to not only be secure, we’ve got to give technology that makes it easier for you, our customers, to secure access to your networks. The technologies that we have built-in and enhanced we think can help you cut the cost of securing the rest of your network through better identity management, et cetera, quite considerably.

What I would like to do now to help you understand this notion of IT efficiency improvements is to show you a video that one of our customers who has been involved with the product agreed to produce with us, this is the Commonwealth of Kentucky, Department of Education, who is significantly improved their efficiency in IT through the use of Windows Server 2003. If you could just roll the video?

(Video segment.)

STEVE BALLMER: There’s no better way, I think, to help you understand the kinds of benefits that come from new products than to actually have the customer who has really put the product through its paces talk to you a little bit about that. And I certainly feel very confident that people who take a look at this product because of how it helps you do more on this one by essentially needing to invest less in the ongoing maintenance of systems inside the data center.

The second area that I talked about of innovation is in terms of improving information workers’ productivity. When we say information worker, that’s sort of a way of saying almost anybody who depends and uses information vitally in what they do for a living. That’s most people, frankly, in most businesses, large and small, around the world: people who want to collaborate, people who want to use e-mail, people who want to share information, people who need to somehow share and analyze and communicate information with one another. Windows Server 2003 is an important product in terms of what IT will be able to do and can do to improve the productivity of the end users on this front.

If you take a look at, today, the marketplace at large, you say, what is the number one market, what’s the number one application of servers numerically in the world? File servers, file servers is the number one sort of volume price that servers get consumed. It’s not 90 percent of the servers, but it’s absolutely the number one consumption place for servers. And any major server release has got to bring something new to that constituency.

What we have in Windows Server 2003 is a new technology, it will actually ship after the Windows Server 2003 product itself, but it’s an add-on that all customers will receive who purchase Windows Server 2003, to download, if you will, to run on their systems. And it provides a new technology that we call Windows SharePoint Services. These services were designed essentially to take the notion of file sharing and information sharing and collaboration to the next level, enhancing the way people from the desktop manage information, letting people communicate and collaborate on information in a different and richer way, but still allowing IT to provide the kind of access and security and protection of information that’s so important. Very big advances, also a set of big advances that you’ll see around the security of documents, e-mail using the new digital rights management technologies that will come out, also, on top of Windows Server 2003.

We think of this as very important, because at the end of the day, the easier it is for people to find and share information off of the Windows Server. That’s something that’s going to bring benefit in terms of more end users than almost anything else that we can do in a product like this. We want to do two things on this front: we’re going to show it to you here live, and we’re going to hear from a customer as well about this. Let’s first do that. And here is how Honeywell ATS, the second largest division inside Honeywell, is using the new Windows Server 2003 and the SharePoint Services Technology to improve the productivity of all information workers inside Honeywell ATS. Roll the video please.

(Video segment.)

STEVE BALLMER: Why don’t we take a look at what the future of file sharing and collaboration really looks like and to do that I’d like to invite on stage with me Katy Hunter from our Windows Marketing Group and we’ll do a little demonstration of Windows Server 2003 and its Information Worker Collaboration Facility. Katy.

KATY HUNTER: Thanks, Steve.

All right, I’m here to talk about how Windows Server 2003 increases the productivity for the information workers that are connected to the server while at the same time reduces the cost for IT professionals supporting those users.

Now, at the foundation it all starts with data availability. In the U.S. alone, companies are spending over $12 billion each year recovering lost data and over a third of that data loss is coming from end users making mistakes.

So we’ve taken on that challenge with Shadow Copy for shared folders, a Windows Server storage innovation that enables end users to recover from their mistakes.

So to show you what I mean I’m going to go ahead and simulate a mistake.

STEVE BALLMER: You never make any.

KATY HUNTER: No. That’s actually why I got cast for this job because I’m so good at making mistakes.

All right, so how many times has something like this happened to you? To create a new presentation you start off with an existing presentation that has one slide that you want to use and you delete all the other slides. Okay, here comes the mistake. Instead of choosing File, Save As to start that new file, you by reflect hit the Save button and there you’ve done it. Forget it about the new file you were going to create; you’ve completely wiped out this original content. It is all gone.

So your options at this point are to rebuild the file or to call your IT staff to see if maybe they can recover it from a tape backup.

STEVE BALLMER: Because this was up on the server?

KATY HUNTER: Right. Hopefully it’s on the server. This is on the server.

So if we could just go back in time —

STEVE BALLMER: Why don’t I just go in the crash bin and get the old one? I’m teasing you; go ahead. (Laughter.)

KATY HUNTER: Okay. So if we just can go back in time and Windows Server 2003 enables us to do so, to recover previous versions of the file right here from our desktop without calling IT.

It’s simple. You just go to the file properties and you’ll see there’s a new Previous Versions tab. You go ahead and restore the most recent copy and you’ll see that when we try to open this file again we’re back on track and the cost of that user mistake has been completely avoided. (Cheers, applause.)

So we have early adopters of this technology and one of them, the Austrian Ministry of Interior, is saying that this feature alone will save them over $2.7 million each year based on lost productivity.

All right, so intelligent file storage is going to save your companies a lot of money but there are even greater gains to be had by enabling team collaboration like you saw at Honeywell with Windows SharePoint Services.

Here we have a Windows SharePoint Services site that’s been created for a virtual team working on a marketing project. So right away you see this is way more than just file storage; it’s a Web-based team collaboration environment that enables all the team members to have a single location for everything related to the project.

STEVE BALLMER: But these are still the shared files, right?

KATY HUNTER: These are shared documents on the server.

So these are shared documents, and it shows who’s working on the document at a given time. It has the upcoming team events and it has the tasks that are associated with the project and which team member is assigned to that task. Here we have important links on the Web for the teams.

And Windows SharePoint Services is customizable so any information that’s important to the project can be incorporated into this site.

Now, here we have the online status of the team members so you can go ahead and kick off an instant message right here from the home page.

So once you get a taste of this new way of working together you won’t want to run another project without it, but those of you in IT don’t have to worry about supporting all of these projects because Windows SharePoint Services is essentially self-service. You can control how it gets deployed and used but each individual team site gets set up and managed and even retired without any further IT involvement.

Okay, I’ll show you my favorite thing about Windows SharePoint Services and that’s how through the support of XML Web services we’ve enabled the team environment to be connected to end user productivity tools like the upcoming Office 2003.

So I’m going to go ahead and open this file and edit it in Word 2003 and you’ll see how my team environment comes along with me while I’m working on the document. So I have the important information here in my shared workspace from my team site. You see the status of the document, that it’s currently checked out to me. I have that online status of my other team members and there’s the tasks, those shared documents from the site and the important Web links; all available to me where it matters the most in the tools that I use every day.

Okay, we have one more thing to show about how Windows Server 2003 increases productivity for companies and that’s through built-in media services. We have here from the team site a link to Windows media. These are live, executive broadcasts and employee trainings that have been archived and made available on demand so the team can view them when it’s convenient for them to get the most benefit out of watching and companies can save the cost associated with live training and live events.

So I want you to notice when I run this video file, what you don’t see and that’s any delays whatsoever, because with Windows Media Services included in Windows Server 2003 and the Windows Media Player 9 series, they’ve virtually eliminated buffering. And those of you in IT can actually manage and throttle the stream, which means rich media will not overtax your network bandwidth.

All right, so we’ve seen how from file storage through to team collaboration and communication services how Windows Server 2003 increases the productivity for information workers while decreasing IT costs.

Thanks.

STEVE BALLMER: Thanks, Katy. (Applause.)

At Microsoft these days we make a point of, as we like to say, eating our own dog food and so we’ve encouraged people a little bit, come on, let’s get some Windows SharePoint Services sites out there, let’s take advantage of the new technology, blah, blah, blah, blah, blah.

Sometimes things happen right away. Sometimes they happen with a little encouragement. Sometimes they surprise us.

We already have inside Microsoft I think over 12,000 Windows SharePoint Services sites build on Windows Server 2003. Nobody sets up a new file share anymore. People are always saying, look, why should I do a conventional old file share? I can sell people so much more about the documents that are here, why they’re here, what we want to do with them than I ever could any other way.

We’ve got a thing now that people created inside Microsoft called My Sites where I can keep my personal storage, I can keep that private but I can also share information about myself, a little bit about my interests, my hobbies. It’s my home storage, if you will. It’s everything to me.

The future, if you will, of the way we store things is richer underlying technologies for recovery, but perhaps most importantly with better user interface and better access from within all the applications that people use inside companies as individual end users.

I’m very turned on by this as a user, as you can probably tell, and certainly our friends at Honeywell and I hope you after seeing Katy’s demonstration share some of that enthusiasm.

The third tag I talked about was essentially delivering applications faster and with higher performance. And between the work we’ve done integrating our .NET runtime and IIS deeply into Windows Server 2003, the work we’ve done improving Visual Studio .NET and the way in which the two products work together we think we have absolutely the highest productivity at development platform in the world. We think we’ve got the best tools in terms of rapid application development. There’s a huge community of people who understand the Visual Studio toolset and who can go build applications.

We’ve made some improvements but we’ve talked a lot about our roadmap to improve the connection between the way applications are built using Visual Studio and the way they later get deployed and operated on top of Windows Server. We have a whole extensive roadmap that essentially starts this process of letting the developer describe applications in such a way that they’re easier to deploy and manage for the IT person.

And last but certainly not least I think it’s clear to everybody we live in an environment today, given the cost pressures, where a lot of companies are looking to migrate off of legacy mainframe systems and particularly legacy UNIX systems and come down to Intel hardware based systems.

And we’ve put a lot of effort into ensuring between Windows Server, the tools that we provide around it and the work we do in Visual Studio .NET that it is very easy to migrate an application from Solaris or from AIX down to the Windows platform.

I think we still have work to do in this area but we’ve made huge progress and I think as that trend just accelerates we certainly have made it a priority to be the easiest platform — that means easier than any of our competitors on Intel hardware — in terms of the migration of UNIX applications to Windows and Intel platforms.

If you take a look at the actual numbers in terms of doing more with less, what we find is that you can write an application. Here we’ve picked sort of an obvious example, which is the Java Pet Store, which is the showcase Java, J2EE application. You can write that application, exactly the same functionality with about one-sixth the amount of code, which is a good indication of about one-sixth the amount of work, and get about twice the performance that you can get on J2EE. And this is the showcase flagship application that Sun has created for the Java environment.

That is what we mean by doing more with less, more application capabilities delivered at lower total cost and with better overall performance, which means lower total cost in the data center.

Rather than have me just talk about it, again we want to have you hear from some of our customers who are benefiting from Windows Server 2003 and what it means from a development perspective so we’ll roll the video now from one of our great customers in England, the London stock exchange, which is the second largest stock market in the world, and we’ll have a chance to hear about how they’re using Windows Server 2003.

(Video segment.)

(Applause.)

STEVE BALLMER: I’d say the work that’s going on at the London stock exchange is some of the most interesting that I have had a chance to see and be involved with and I think you can certainly appreciate some of the benefits that they’re deriving from Windows Server 2003.

We thought we’d show you a little bit of this productivity at work, doing more with less in terms of application development, and to do that Prashant Sridharan from our Visual Studio .NET group will join us here for a little bit of a demonstration. Prashant.

PRASHANT SRIDHARAN: Thanks, Steve.

So as you saw, the London stock exchange used Windows Server 2003 and Visual Studio .NET 2003 to deploy applications faster than traditional application servers and build a whole new class of solutions they could never have built before. I’m going to show you a demo of a kind of application with the kind of technology the exchange used to achieve this.

Now, historically you’ve had to choose between making your developers more productive and helping your IT staff maintain a reliable infrastructure. I’ll show you how productively your developers can now work but more importantly how you can do so in a way that sets your IT staff up for success and deploying and operating those applications.

On the big screen behind me you can see some real time security status, as represented by the information on the stock ticker. This data feed is sent to a server application, which enriches it with complex financial calculations and is then sent to several thousand trader terminals like this one.

Now, many of your applications require absolute performance, and in this scenario the enhanced data has to pass through the server application and show up on the trading terminals in under one second.

But high performance isn’t the only requirement. You also need your IT staff to be able to monitor your app at all times and enabling application manageability is a core feature of the Microsoft application platform.

What I’ll do now is open up the server application in Visual Studio and add features to assist in the deployment and operations of that solution.

Within Visual Studio, I can use the server explorer to gain quick access to features and components of Windows Server. Using the server explorer I’ll instrument my application with Windows built-in performance counters. This simple drag and drop operations saves hundreds of lines of code and lets you take advantage of the operations features of Windows Server 2003 from within the Visual Studio environment. I could then write code to use those performance counters.

Now the app is modified. So before I send the modified app to my IT staff to test I’ll first set the version number so they can manage it more effectively. Changing only one line of code, you can version your solutions so they’re easier to deploy, operate and if necessary roll back.

With this as well as IIS 6 features such as process recycling and kernel mode caching you can quickly and reliably deploy your app without rebooting your server. And with a server app that includes instrumentation your IT staff can monitor its performance using any WMI enabled monitoring solution such as OpenView, Tivoli or Microsoft Operations Manager or a custom monitoring application like this one.

Now look at the performance monitor. You can see the counters I added to instrument my code across the top and along the bottom you can also see that with the built-in load balancing features of Windows Server 2003 our infrastructure scales to meet the sub-second performance requirement of our application even if the server load is high. In so doing you can scale your systems based on the needs of your applications.

Now, one of the things you’ll find when you have more efficient operations is that you have the time and money to offer more value-added services and focus on the business value of your app without worrying about plumbing.

In between, when the data comes to our server and when it’s send to the trading terminal, we happen to store that information in a SQL Server database.

Because we now have the time and money to do so, we can monetize this information. So to that end let’s also offer customers historical data on how the market performs. Because we already store this data it’s now just a matter of making this information readily available. We’ll do this by creating a Web service so that customers can pay us to use it in applications of their own.

In this scenario you can see that our customer, Woodgrove Bank, is building the Web site and now needs to connect it to the historical data in our database. To do this they can use the integrated UDDI features of both Windows Server 2003 and Visual Studio .NET 2003 to search for our Web service. In this case they’re looking for the historical data service and once they find it they can select it and consume its data inside their Web site. Since the data is returned in industry standard XML they can easily integrate it with any system.

Let’s view the customer’s ASP .NET Web site now. As you can see, all the historical data stored within my data warehouse is incorporated into my customer’s Web site and if you look at the bottom of the page you’ll see that they’ve also included market information provided by other Web services vendors.

Now you’ve seen how the Microsoft application platform can meet the rigorous requirements of the most demanding enterprise class applications, including the London stock exchange. Through built-in instrumentation and versioning capabilities the Microsoft application platform prevents problems typically associated with deploying and operating solutions, and with native Web services support the Microsoft application platform lets you connect with your customers, partners and existing investments. And with an award-winning developer tool that unlocks the power of the platform programmers have never been more productive.

With Windows Server 2003 and Visual Studio .NET 2003 as your application platform you can deliver solutions faster with less code and maintain the rock solid reliability your business demands.

Thank you.

STEVE BALLMER: Thanks, Prashant. Yeah, nice job. Thank you. (Applause.)

The depth in what Prashant just showed you is really, in my opinion, quite amazing. When he talked earlier about the built-in UDDI services in Windows Server 2003, yet they were fundamental to the kinds of productivity that Prashant showed you just there in extending and building off of the core trading applications from the London stock exchange. Again, I remind you the product is deep and it has a lot of capabilities to help you do more with less.

I want to turn now to the issue of scalability and I’m going to talk about it a little bit. We’re going to hear from customers about it and then we’re going to have a chance to do a little announcement.

The first thing I want to talk about is some of the amazing results that we have been able to achieve with our ISV partners, in this case particularly Siebel. Siebel is one of the most important enterprise applications in the world today. We at Microsoft internally use Siebel quite extensively. It is important to us as a company to prove enterprise readiness to make sure that the key applications can run at maximum performance on our platform.

Because of some good work done at Siebel, good work done at Microsoft in our SQL Server 64-bit edition and in Windows Server 2003, all running on Intel’s Itanium platform, we now have the number one high-performance benchmark for the Siebel application, 30,000 concurrent users on a single server from Unisys, running Itanium, ranked number one. There is no hardware/software combination you can buy from Solaris to AIX to mainframes, there is no system in the world that can handle more concurrent Siebel users than a Windows Server running SQL Server sitting on top of a Unisys system running on Itanium architecture.

It really is an amazing, amazing thing to stop and think about if we only go back four or five or ten years when people were still wondering will the PC server platforms ever be able to run the biggest jobs in the biggest businesses in the world; number one, number one top performing benchmark, absolutely, of any platform in the world.

To help you understand a little bit more about this from somebody who’s really putting Windows Server 2003 through its paces in a high scale environment every day I’d like to show you a video made by one of our customers, JetBlue, talking about how they use Windows Server and Windows Server 2003 to scale up and run the mission critical applications at this airline. Roll the video, please.

(Video segment.)

(Applause.)

STEVE BALLMER: I hope everybody there was applauding the Cohen family. I don’t know if you noticed but both the CIO and the chief architect shares the same name and certainly we’re glad for the support from JetBlue and the incredible work that Jeff and Adam have both done.

Itanium has been an important milestone I think for the industry. I talked earlier about how we’ve been working 15 years essentially to scale, to solve the biggest jobs in the biggest enterprises in the world. We saw the Siebel number and the Siebel benchmark. It has taken advances in the software, in the chips and in the systems to make that kind of performance happen.

A partnership that has been absolutely critical for us on that is the incredible partnership that we have with Intel. People associate that with the work that the two companies have done on the desktop but I’d say in some senses the pace of advance has been even more rapid in the server environment over the last several years.

I promised you earlier a little bit of news on this front and to share some of that news and give you a perspective on where we’re going for scalability I’d like you to please join me in welcoming the Chief Operating Officer for the Intel Corporation, Mr. Paul Otellini.

(Applause.)

PAUL OTELLINI: Thanks, Steve.

As Steve said, Intel and Microsoft have had 20 years now of a spectacular partnership, and you’re right that for most of that time people associate us with delivering year after year better PCs. But for the most part of the last 10 or 12 years we’ve really focused more and more of our energies on the server and I want to talk a little bit about what we’ve accomplished in servers.

Over that period of time we’ve been able to drive the absolute performance leadership in two-way servers and four-way servers and single processor servers, in clustered machines, and given our customer’s great value, as you said, of doing more for less by driving the cost per transaction down every time.

But you know, Steve, there’s been one part of the market that’s just eluded both of us and that’s really the absolute performance crown and I wanted to talk a little bit about that today.

You know, we’ve made some nice strides the last few years together and on this slide it shows where we are together. In 2001, with Itanium we got all the way to number six. Now, none of us settle for number six, as you know. This is against the absolute highest performing, non-clustered machines that are out there. The green bar on the slide is the fastest RIS machine with transactions out there today. It happens to be a Fujitsu machine based upon the SPARC microprocessor. It’s a 128-processor machine, a big machine, mainframe class machine.

Last year with the Itanium II we were able to get all the way up the number five position. Okay but still not great.

But, you know, as of last Monday using an Itanium II machine from NEC — it was a 32-way machine — we got all the way to number two.

STEVE BALLMER: Number two.

PAUL OTELLINI: Right, but not good enough and we’re trying harder.

I am really happy today to announce something that was announced in Japan yesterday morning, which is an NEC machine, that same NEC Itanium II machine was upgraded with a new chip from Intel, code-named Madison, which is the next generation of Itanium II, simply plugs into the machine. It gives them a performance boost and their TPC numbers today on that same 32-way machine have given them the number one non-clustered performance, given us the non-clustered performance championship in the world. So we did a really good job here. (Applause.) This is, of course, with the Madison chip, with Windows Server 2003.

Now, the interesting thing about our architecture is we continue to scale, right? And one interesting thing I wanted to throw out into the audience today is another announcement. This morning Hewlett-Packard is announcing that they’re now number one. (Laughter, applause.) And this is an awesome machine. In fact, guess what, this is the machine. It’s their Superdome machine. It’s a 64-processor machine that also has the Madison version of Itanium II inside it. It’s running Windows Server 2003 and SQL, Microsoft SQL. So we’ve now gotten the TPCs up to 658,000, so absolute record machine.

STEVE BALLMER: You mean there is no machine in the world that could do more database transactions than this machine running Windows?

PAUL OTELLINI: Yeah, this is the absolute fastest transaction machine on the planet. (Cheers, applause.) Finally. Finally.

Now, those of you who know Intel and Microsoft know that being number one is a good place to be but you have to fight to stay there, and we’re not finished. This journey is going to continue and I’d like to maybe share a little bit about where we’re going here.

The chart here shows transactions and the top dot on the chart is really based upon what is inside the Superdome today. In the next 18 months there’s going to be some very exciting products from Intel. The Madison chip will go into production. We’ll launch it mid this year. And the second half of this year we’ll launch a product that’s codenamed Deerfield, another member of the Itanium family, which is aimed at lower powered systems optimized for two-way and workstation environments, so taking the architecture down.

But we don’t want to take our eyes off the high-end performance. Over the next 24 months you’ll see us bring out another version of the chip with much larger cache using a new silicon technology and beyond that in 2005 yet another version of the chip using multiple cores. These two chips will take the absolute performance up another two to three times where this machine is today.

And we aren’t stopping there; we also have in development our first billion transistor chip microprocessor that is another member of the Itanium family that will come out in a few more years that will take the performance up by a factor of ten – a factor of ten from where this is today. [We’re] fully staffed, well into development and we’re well on our way to delivering that. So we haven’t stopped driving performance in terms of scalability.

There’s an interesting thing about the architecture I wanted to point out though, that we’ve built in not just the ability to scale up in terms of performance or number of processors but the ability to address more and more memory. With these machines we can now address over a terabyte of memory inside of a single machine. That’s architecture.

We’re also taking the machines and improving them relative to user visible benefits. The first of those is to allow the architecture that’s inside of Itanium, which is a very parallel architecture, to be available to your SQL team to take advantage of that parallelism and drive database parallel task performance up substantially, which takes the number of concurrent users up dramatically again, user benefits for our collective customers.

But I think the thing that’s most important is the reliability. You pointed it out earlier in your demo. The Itanium chip includes an advance machine check architecture that provides all those hooks for your operating system to be able to allow the end user, the systems administrators to be able to see what’s wrong, avert disasters and recover from disasters very, very quickly. So we’re dealing with this at the architectural level as well as at the silicon level, and I think that we are delivering very interesting value.

The scale that we’re operating on together, the scale that we’re delivering is having a scale effect on our customers, and I thought I might just point out a little bit in terms of where we are today.

The first Itanium chip that came out had ten OEM systems running in the two and four-way and only two customers eight and above. That was pretty good. That was our first foray into this space. In 2002 we more than doubled that, over 20 customers at two and four-way running systems and we’ve taken the number of eight-way and above systems up to five. Better.

This year with products based upon Madison we are really moving into the mainstream here. There will be over 40 OEM systems running two and four-way configurations and over ten sitting here at eight and above.

And I thought I would just spend a minute and maybe do a little walk through of what we have up here. This is the HP Superdome machine we talked about before, 64-processor machine, the world champion. This is the epitome of scaling up in my mind.

This machine is from Unisys. It’s a very interesting machine. It has multiple partition capability. Inside of this machine you can have a 32-way Zeon, 32-bit partition and two 16-way Itanium partitions all having the same software transparency and homogeneity, providing our customers even more value, sort of a mid point.

STEVE BALLMER: It’s very important, I’ll point out, for example, that JetBlue is using this machine.

PAUL OTELLINI: Exactly. Exactly this.

And then on the other end, on the scale out side of the business this is a Dell machine that happens to be an eight-server machine that’s running in a clustered environment. Each server is two Itanium II processors and it represents sort of the best of scale-out today that’s in the industry.

So we are now addressing best of class scale out, best of class scale up and everything in between with our products, and that’s why the customers I think are lining up.

The last point I wanted to make is really where we think we’re going in the next two years, in ’04 and ’05. We have together design wins now that will more than double the machines again, over 80 OEM systems in the two and four-way and over 15 in the eight-way and above, and we’ll continue to bring world class results here.

STEVE BALLMER: Incredible. Doing more with less, there’s never been a better example than Itanium. Thanks, Paul.

PAUL OTELLINI: Thanks, Steve. (Applause.)

STEVE BALLMER: It blows my mind. I mean, quite frankly it blows my mind. When we originally hired Dave Cutler, who is the architect of our NT systems, you know, I was trying to convince him to come to Microsoft and he said, “Look, I don’t want to work on toy operating systems. I want to work on real operating systems.” It may have taken us and our partners at Intel and others a while to get there but with this amazing result that Hewlett-Packard has posted today with Itanium I know I can now look Dave in the eye and say we’re past that for sure; no more toy operating systems, the highest performing system in the world. (Applause.)

I want to turn now though from the biggest of the big down to smaller business. If you take a look at where servers get used in the world today, about two-thirds of them get used in small and mid-sized companies. And we need to make sure that we’ve got a product and a technology that scales not only up for those very large jobs in the biggest enterprises, but also down for the simplicity and ease of use that’s necessary in smaller organizations.

The first person I want to have you hear from on this is a company called Fleischer’s Bagel. Let me take a small show of hands: How many people had a bagel before you came in here today? Bagel, anybody have some bagels? Those were Fleischer’s bagels that you had this morning and Fleischer’s Bagels is a good user of Windows Servers and we’re going to hear a little bit from them. They’re a good sized company but the IT staff is essentially one person who also has other operational responsibilities and everything else they do they do with outside vendors. So let’s hear a little bit how Fleischer’s Bagels is benefiting from Windows Server 2003.

(Video segment.)

(Applause.)

STEVE BALLMER: When it comes to making bagels or just eat a bagel in the middle of the day I’m glad to see Windows Server 2003 is helping smaller companies also do more with less.

I’d like to invite on stage now with me Mike Iem from our marketing organization who is going to show you a little bit some of the kinds of things that you can do in small business environments with Windows Server. We will be bringing out here subsequently the Small Business Server Edition of Windows Server. It will be out just in a few months and Mike’s going to give you a peek at some of the kinds of things that you can do in small businesses. Mike.

Iem: Thanks, Steve.

As you’ve just seen in the video, the simplicity innovations in Windows Server 2003 allow Eric at Fleischer Bagels to make it easier. As you’ve just seen in the video, what I’m going to show you is how easy it is for Eric and other part-time IT pros to manage Windows Server 2003 with the manager server tool, a tool designed to bring into one location all the key server management tasks.

From this single view you can manage all the roles that the server is providing. You can see that we’ve enabled several roles on the server, each with their associated management tasks. It’s simple to add and remove roles to the server. Add or remove roles reduces potential errors and saves time.

So when I click on this link it will launch the Configure Your Server wizard.

So let’s go ahead and see how easy it is to set up a file server on this server. So I’ll go ahead and take the defaults and I’ll go ahead and give a location where the users can store their files and I’ll set up access security and it will help me do that, so that users have read and write access. You can see that the wizard does all the work. And I know that I’m not going to miss any steps because it’s unified into a single wizard.

And we’re done. And now we’ve just added a file server to this role and only those services necessary to run that role are enabled and no others. It’s that simple to set up, making for a reliable and a secure system.

Now when you need to manage the role over time, you come back to the same location and the same place and present it with just the management tools you need relevant to that role.

So let’s click on the manager file server role and mange the file server and let’s say that we might want to know what users are connected to the server and what files they have open.

So Windows Server 2003’s simplicity means less effort and less time are required to keep the server running smoothly.

So the next wave in simplicity innovation comes in the form of Windows Small Business Server 2003, and one of the most demanding requests by users, one of the most difficult for IT, small business IT to implement is configuring and connecting remote users to the network of the server.

So with Small Business Server 2003 we made it easy. Employees just go to one company Web site, securely from any PC with a browser, from anywhere in the Internet and simply log in. And you can see here I just went to my NorthwindTraders.com site and I’ve logged in and I’m presented with these connection options, so I can read my company e-mail, I can access my internal Web site or I can connect to my computer at work. It’s easy for users because it’s a single place to go and it’s easy for the IT person because it’s set up right out of the box.

So how many times have you wanted to get access to important information on your desktop at work but you weren’t at work? So with the wizard here we just click on Connect to My Computer at Work and based on your logon information it knows what your computer name is and when I click this button it’s going to connect me right back to my PC at work and I have access to all my important files. That’s how simple it is.

So I showed you how simple it is to configure and manage server roles, how Windows Small Business Server 2003 is further delivering a simplified server experience so that small business customers can get the great business returns from server-based computing.

Thank you. Thanks, Steve. (Applause.)

STEVE BALLMER: Thanks.

Windows Server 2003 is a great product but it is part of a family of products that we think are quite important for our customers, the Small Business Server that Mike just showed you and a variety of other products that add utility for our information technology professional customers and developers, security products, management products, application development database products, e-mail, et cetera, all part of what we call the Windows Server System, a family of applications targeted and helping people do more with less.

I talked at the beginning about how excited I was about the level of interest and enthusiasm we’ve seen from customers. The same is true in terms of what we’ve seen from the industry. There are over 220 top applications ready today that run on Windows Server. There will be over 2,500 applications within 180 days. We have a number of partners trained. There’s systems built. There’s applications. You’ve seen the level of enthusiasm from Unisys, from Dell, from Hewlett-Packard, from Intel; all of which leaves me feeling confident that Windows Server not only is something that will benefit our customers and benefit Microsoft but will be a real boom in terms of productivity and efficiency for our industry partners.

I get asked a lot how does Windows Server 2003, where does it fit for you competitively. I talked a lot this morning about the enhancements and the changes and what we’re doing and the new stuff, but at the end of the day we need to make sure that people are convinced they can do the most with the least with Windows Server 2003. That’s what everybody in this audience will always need to evaluate.

Clients are a little different than servers. For clients you want to have all those applications running on a single machine. In the server environment there tends to be more focus, particularly in large companies, of machines on single applications.

This is a list of some of the things, quite a few of the things that our customers try to do in the server environment. They want to build applications. They want to extend legacy applications, port applications, develop new ones, database, Web pages, high-end, high-performance computing systems, scaling up and out, information worker infrastructure, business intelligence, portal, workflow, collab, IT infrastructure. We think in most of these scenarios we have absolutely a compelling comparative proposition for our customers.

We love competition. We love it. It pushes us to innovate and do the best work we can for our customers. And in so many of these areas we feel like our customers will agree, as JetBlue, as the State of Kentucky has and others that we are way in front.

In some of these scenarios we know we still have some work cut out for us and you’ll see us push and innovate and integrate new technologies to make sure that no matter what the scenario, no matter what you’re trying to do in your business we help you do more with less; in fact, we help you do the most with the least.

And with the advances in reliability, security and manageability built into this product we think we’ve absolutely got the rock solid foundation that should give you maximum confidence in running your business on this platform.

We think that the key is to really push forward on all of these aspects. We’ve got to follow some principles. We’ve got to innovate and integrate new technologies. We’ve got to be strongly interoperable and our XML story is good there. We’ve got to make sure we’ve got the best tools for developers. We’ve got to reach out broadly through events like this in the community and we have to offer simple high value customer experiences.

And if we do all of that we think you will agree with us, as many customers as you’ve seen today have, that we have absolutely the right products for the right times to let you do more for less.

Thank you very much. Go enjoy the pavilion. We’ve enjoyed it a lot. Appreciate it.

(Applause.)