State of the Union. It's been one year since I last stood up here and talked about how Apple is doing in the enterprise. And boy, it's been an incredible year. We've had an increasing product line aimed at the enterprise. We've had very aggressive adoption by our partners and some great feedback from our customers and from the industry in general. To start out, I'd just like to give you a little feel for that. Here's a quote from P.J. Conloy in InfoWorld. And this touches on a very interesting subject. The combination of the cluster node configuration and X-Grid clustering software represents a revolution in how IT looks at computers and builds systems. And if those of you who remember last year, I mentioned that we were going to first see clustering in academia, but the very same things that were driving clustering and its use in research were going to find their way into IT. And Apple jumped ahead of the curve there and jumped right into clustering in academia and science. And guess what? We're starting to see it in the IT shops as we speak.

I'll talk a little bit more about that as we go on. Another quote from Tom Yager at InfoWorld. If you want a complete productivity platform, you can nickel and dime your way there with Windows. Actually, I think nickels and dimes are probably too small a denomination for that. You can hammer and saw your way there with Linux. And for anyone who has deployed Linux in the enterprise, I'm sure you know what I mean by hammers and saws or what Tom means by hammers and saws. Or you can hit the ground running with OS X. And that is really the key differentiator. I mean, OS X, if you look at it, is chock full of open source software, runs a lot of open source tools, as you would in Linux. But the big difference is that Apple, we take the responsibility for integration, for QA, for testing, for security updates. We give you a way to hit the ground running So it's kind of the best of both worlds. You get the support you would get from a proprietary system, but the full power of taking advantage of the world's ability to create open source software with all of that available on OS X. Just an incredibly increasing number of enterprise customers over the last year actually in deployment. And we're going to actually hear from some of those guys later today, they'll kind of take you through their decision process and where they are in that process. I think you'll find that very interesting.

So innovation, you heard about this from Steve this morning. If Apple is anything, Apple is an innovative company. I think we almost stand for innovation in our industry. And you hear about innovation a lot in terms of the consumer side of our business. So the iPod or the iTunes music store and what we've done to the music industry in general. or iLife and the consumer products that we make available.

However, innovation also translates into the world of IT. And in the world of IT, it's not just innovation around ease of use, although that is important because ease of use translates directly into less calls to the help desk, et cetera, et cetera. But innovation in the IT world also translates into ease of ease of management ultimately translates into lower total cost of ownership and that's where we focus our innovation and our skills at innovation in the IT world both in the software and in the hardware now I'm a software guy so I'm going to spend a little bit of time talking about the software side and innovation here is is focused on OS X. And the latest innovation there is, of course, OS X Tiger.

But this is really the culmination of a number of years of Apple moving increasingly into the IT space. Starting back in the year 2001 is when we came out with our first OS X server. We're on our fifth release of OS X server. And this is OS X repackaged with the things you would expect in a server environment aimed at the IT marketplace. It's actually built off the exact same OS X Unix-based source code as our client products. That's a huge advantage for us internally, but also for you because any innovation or performance increase or bug fix or security fix goes into that single source code base. And that line has continued starting from 2001 up through OS X Tiger Server, which is our most recent version. On the hardware side, in 2002 we introduced XServe, so Apple's entry into the 1U dual processor server market. That was a very successful entry for us. It kind of caught people's attention because they said, "Gee, Apple makes a rack mount 1U product. What's and broadened this product line over time. Also in 2002 was our first storage product, XServe RAID, which is a 3U rack mount unit with full RAID storage. And as you see, as the years progressed, we continue to add to and complement and increase this product line, not just with software and hardware, but also with service and support products products designed to meet the needs of the enterprise. Towards the end of 2003, this is when we sort of caught or it caught our attention that cluster computing and grid computing was going to be a big deal, not just for the scientific researchers but ultimately for IT as well. One of the first places that this showed up was in the biotech markets, bioIT, and we We came out with our work group cluster server for BioIT, which was a big hit with pharmaceutical companies, companies doing genetic research, etc. So this trend has continued on through 2004, 2005, with increasing, not only increases to our server and storage product lines, but on the software side, some very innovative things.

I want to just mention a couple of them. Apple Remote Desktop, a management tool for managing Apple desktops, really an incredible set of features ranging from how do you deploy software to how do you remotely manage that desktop to how do you provide a help desk service to that desktop. I'll talk a little bit more about that later. The other thing I want to mention is our SAN product line. We have XSAN, which is a fully featured storage area network.

Our customers, as any IT customers, kind of the first word on their lips is storage. There's an insatiable need for storage, and the best way to manage storage, if you have a lot of it, is with a storage area network or a SAN system. It allows you to add storage over time without reformatting, very flexibly managing your volumes, having backup solutions that are industry compatible, etc., etc. In January 2005, this year, we introduced the XSERV G5, very high powered in terms of processing speed but equally important and anyone who knows about architecture and balanced architectures this has a dual 1.15 gigahertz system bus that provides a total bus throughput of 9.2 gigabytes per second bandwidth per processor so for a dual processor system you you double that that is incredibly important for any data intensive operations which pretty much any IT or transaction based operation is. So, you know, processor is part of the equation here, but equally important is Apple's innovation in terms of how we manage the bus, how we hook these pieces together in such a way that we get maximal throughput all the way from I/O to processing to the the intercommunication that has to happen between the processors.

XSERV RAID, again, storage is an interesting story because the rate, if you were to plot the rate of storage available at a certain price, that has been exceeding Moore's Law for quite some time now. Moore's Law says you double every 18 months. Storage has been pretty closely following a 12-month doubling time, and that shows really no sign of slowing down. And that's actually matched or exceeded by the need for storage in the IT environment.

So Apple's XServe RAID is designed to meet this need. We've been very innovative about and smart about picking drives that can deliver basically the greatest number of bytes per buck that anyone in the industry can deliver in such away that it's industry standard, fiber connected, fully rated, so it's highly robust and redundant. Our current price is $2.32 per gigabyte. If there's one message you take away from this, don't do a procurement and storage without least looking at Apple XServe RAID. It's an incredible value. And Apple is committed to just following this curve as it progresses. And believe me, your storage needs will continue to increase.

Mac OS X server, this is Tiger server. What's new in Tiger server? I'm going to actually go into this in a little bit more detail later on. But just to mention the key aspects that are new, 64 bit applications, that means 64 bit address space so that you can go beyond the 4 gigabyte barrier. ACLs or access control lists, big deal in the enterprise. iChat server, run your own instant messaging inside the enterprise securely. Software update server gives you control over when systems get updated inside the firewall or inside your enterprise. Web log server, web logs are essentially the groupware of the future. You're going to, if you're not already running them inside your company, you're going to be running them inside your company. We integrate a blog server right in with Mac OS X server. Junk mail and virus filtering, which is a must for any mail system that's deployed by an enterprise today, and we have some very good products there. Ethernet link aggregation, that's sort of under the hood, but I'll talk about that a bit. And then Xgrid, taking advantage of those unused CPU cycles around your corporation to do high-performance computing.

I mentioned Apple Remote Desktop. Management made easy. It's software distribution, so a way to easily bring many desktops all in parallel up to the same rev level in terms of software. It does asset management, remote administration, remote assistance. It's based on standards like WebM. For remote desktop, it's based on standards like VNC. So this is a great product for anybody who has to manage Mac desktops. There's hardly anything of this caliber available for other platforms in the industry. It's a great product. I mentioned XAN. The key thing to take home about XAN is that this is not a toy. This is a fully featured, high-performance storage networking system. It is for fiber-connected storage with back-end flexibility for hooking into industry-standard backup solutions, easy remote administration, flexible volume management, so as your storage needs increase, you can just toss more volumes on this or more physical storage on this and not have to reformat, not have to repartition. Anyone who has large storage needs absolutely is moving to sand-based storage, and Apple has a great solution here. If you kind of open up the hood and look one level deeper, look at the price that we offer on this, and you probably, I guarantee you, you will have a hard time finding anything that's going to be more cost-effective.

I mentioned aggressive partner adoption. We've had a huge number of IT and enterprise partners come on over the last year. Just to mention a few, Oracle. We talked about Oracle 10G last year. Oracle 10G is now available. TIBCO, another partner that you're going to hear a bit more about from one of our customers later on. So this momentum continues. People now actually think of us when they think of the IT space. They put us into their procurements in terms of evaluation. I think this momentum is only going to get stronger and continue.

So speaking of customers, a very interesting one is Oracle. Now, Oracle had a need for low-cost storage solutions, both for their internal deployments, but equally important, something to recommend to customers. And Oracle sort of views themselves as providing transaction-based use of storage. So they provide the atomic transactions that make sure that the data that gets transacted absolutely gets there. What they need at the back end is the most cost effective storage solution possible.

So they evaluated just a large number of potential solutions there. And guess what? They decided that one of the best solutions out there was XServe. So not only did they write a white paper and talk about this kind of storage and how XServe can play into that environment, they actually practiced what they preached. And they deployed internally at Oracle a 1,000-seat implementation of Oracle Collaboration Suites. And this has been very successful. They're expanding it to 5,000 seats. And I think it's a great example of a partnership that's kind of blossomed into something beyond what we expected. happy they're selling it to their customers but for them to deploy it internally makes it that much better of an advertisement kind of an endorsement CNN CNN obviously a large company storage again I'll keep coming back to storage CNN has just incredible need for terabytes upon terabytes of storage and And they're a 24/7 operation.

CNN cannot go down. I don't even want to think what would happen if CNN went down. And what they have done is they've deployed 20 XSERVs, six XSERV RAIDs, 32 XSAN file systems, all front-ended by 30 Final Cut Pro desks, where they do their editing, again, 24/7 for all the new shows. the ones you're familiar with, of course, on cable, but also the things you see in the airport, and there are other outlet channels on the web, and so forth. The result is that this is mission critical for CNN. It's a 24/7 operation, and it's managed seamlessly from start to finish in the workflow on Macs, a great example of what we can deliver to this kind of environment. It even stretches back in the workflow out into the field. So for example, out in the field, there's 120 Mac laptops or PowerBooks together with cameras and satellite transmitters. So when you see a field report coming in from CNN, that's actually being delivered via Mac laptop.

Finally, or one more story that's of interest and sort of centers around not just 24/7 availability is the army.mil website. It's actually about 60 different websites that the Army runs and these are deployed on XSERVs and XSERV raids. The interesting thing about this is as you can imagine, these sites get attacked occasionally by hackers. And they've been running these sites on Mac OS X since -- for about the last five years, and the result is no successful attacks.

So not only do they have a site that, you know, just stays up and keeps running, but there have been no successful hacker attacks on this site. The over a terabyte per day this is not just a static website they actually make heavy use of streaming and Apple has a lot of great server-side products for for quick time streaming that's part of their solution and so this is this is a very heavy traffic site it's a heavy amount of bits streaming out of it getting hacked all the time it just keeps on running so a great great of MAX being deployed into a very mission-critical environment.

This one is fun. This is UCLA physics and astronomy department. This is, these examples just keep coming up of clusters getting deployed of max. Clusters of 64, 128, 256, 1024 clusters of max. It turns out to be just a great solution. In this case they're using it, they've got 256 dual processor XSERVs and the result is they can do 4 trillion double precision floating point operations per second.

They use this to do fusion research at UCLA. And one of the most fun things whenever one of these gets deployed is especially in a university environment is when everyone sort of gets together to to assemble the thing. And in this case, they did some time lapse movies and put together a little film showing it kind of getting built from the ground up. And I thought I would just share that with you because it's fun to watch. You want to cue the movie, please? Guys?

where they say it's easy. Last one is, if you were here last year, I talked about COSA. COSA is-- I guess what I can say is they're a federal contractor, and that's probably all I can say. But they have a need for large-scale, high-performance computational clusters for hypersonic research. And that's exactly what they do. They are currently the largest Mac cluster. They have 1,566 dual processor, 2 gig G5 servers, backed up with XSERV and XSAN systems. And they simulate hypersonic flight. I'm not sure exactly what's going on here, but I think this is something you don't actually have happened during hypersonic flight.

But another example of just the rate at which cluster computing is catching on. So year in review, again, a growing enterprise product line. Our product line continues to grow both software, hardware, and support products. Aggressive partner adoption, guys like the developers who come to this conference, continuing to port to Mac OS X. And then finally, some great customer feedback as they begin to deploy into their environments.

Now what I'd like to do is introduce some real live customers to give you in their own words some of their experiences and where they are in deploying Mac solutions. First of all, I'd like to welcome Vijay Sante who's the Chief Information Officer for Broward County School District. Now you don't necessarily maybe think of a school juristic as being like a Fortune 500 CIO position. Vijay has extensive experience as CIO. He was named, actually, CIO of the New Century by Business 2.0 Magazine, has managed multimillion dollar projects. Broward County is like a Fortune 500 enterprise deployment, and he's going to go into some detail on that. So without further ado, I'd like Vijay to come on up and tell you about what they're doing.

Thank you, Bhad. That was very, very interesting. And again, once again, I'm very glad to be here at WWDC to actually share some of the unique challenges the educational enterprise presents and also teaming up with Apple, how we have been very successful in leveraging our business partnership with Apple and are able to meet some of these unique challenges. The role of a CIO in education is very challenging and some of the unique tasks that a CIO is faced with include a very important one, which is how do you leverage this technology and restructure and reform the use of technology in education? How do you effectively integrate technology into the teaching, learning, and research process? And finally, how do you serve the needs of the key constituents, which would include the teachers, the students, the parents, and not to forget the administrators. So like any typical enterprise, we have the same operational challenges, and in addition, we also have our institutional and informational challenges. So all these lead to some very unique process requirements.

As we all know, Broward is a very large school district, and we have our own mission and goal, and our goal is very clear. Our goal is to effectively leverage technology and revolutionize education and also enable student achievement. So using technology, we are trying to enable change, which is giving every student a laptop so that they can adapt this model of 21st century learning, be able to effectively increase the use of technology in schools, and finally be able to integrate all our disparate applications at the enterprise level. And that's what we're shooting for. And what this provides is with Apple technology, we are able to customize and personalize education, and we are able to meet the diverse needs of our students. Thank you.

Again, let's get started with some very interesting facts. As you can see, Broward is a very large organization. We are the sixth largest school district in the U.S. and we have over 274,000 students, an operating budget of $4.2 billion and 39,000 plus employees. So clearly it's the size of a very typical Fortune 500 organization.

So let's look at some other statistics which present the education enterprise. And in an education, we are faced with the same challenges other industries face. Our goal is clearly to create an enterprise architecture that will unify our applications, our databases, streamline our information delivery, and finally create an integrated system. So we are clearly using some of the leading technology. We use SAP for our ERP. We are 80% Apple shop. We have over 150,000 network devices. And the list goes on. I think we have almost every vendor product in Broward. We're working with IBM, SAP, Cisco, Microsoft, and the list goes on.

What you see here is a very interesting slide, and this kind of shows the complexity of the educational enterprise. And mind you, it also resembles a typical Fortune 500 company. What you see here is an ERP system, and if you look in detail, you will see that there are a lot of legacy systems, and these are mission-critical applications. You see decentralized applications and databases. In a nutshell, it's a fragmented technology landscape.

In addition to the challenges enterprises face, we also have our own educational challenges. One key challenge is the No Child Left Behind Act, which Congress had passed in 2002. If I were to compare this act, I would compare it to the quest of Congress in the 1960s to put man on moon. It's a very pressing, demanding law that looks at accountability and by year 2014, the goal is to make sure that we're improving academic performance at all levels for all students and eradicate illiteracy completely by the year 2014. So that poses some unique challenges and we are faced with the need to have very strong testing protocols in place, reporting requirements, and finally accountability. So our goal at Broward is by using Apple's technology to effectively integrate these silos of information and move towards a complete total solution.

To make this happen, we're taking a very systematic approach. We are using a building block approach where we are layering our technology. We're starting at the foundation with infrastructure, and we have deployed Apple Xers in all our schools. That's 251 schools and increasing day by day. We have laid a solid foundation, and our goal is to build advanced applications on top of that and effectively integrate people, process, and technology resources. Our goal is to reach to the top of the pyramid, and that's the knowledge layer.

To get there, to get to the portal layer or the knowledge management environment, we are embracing open standards, and this is where Apple is a good fit. Apple's OS X Tiger is embracing open standards and open technology and works very well with our goal to take our diverse legacy applications and effectively integrate it into our educational enterprise. So here we go. Our goal is Project Nexus. And project nexus is Broward's Enterprise Portal. Our goal is to have available a single gateway to personalize information, which will help our key constituents make the right decisions. We're able to create a unified view of information and finally knit together and streamline all our processes.

What you see here is a very high level schematic, and this is the pain enterprise organizations go through in creating a portal. You start with your bits and pieces of data information. You use ETML tools and create databases and data warehouses, and accessing the operational data source and databases and data warehouses are the key strategic applications, which are all served through a portal. This is how IBM with WebSphere, Microsoft with SharePoint, and SAP with NetWeaver enable portals and are effectively integrating people process data and applications. I think with Apple and widgets, there's a better way. I think we're able to lower the price point by opening up these doorways and creating an intelligent desktop. And what we see here is by embracing widgets, we see a perfect harmony or a perfect symphony between portals, business intelligent tools, and infrastructure. So with widgets, we are able to effectively integrate people and have readily available at our fingertips the access to portals and collaboration, be able to get to the right information, whether it's residing in a data warehouse or as a business intelligence tool. Finally, streamlining our processes. And the key to application integration is Java and J2EE. And that's another open standard which Apple has embraced in their OS model.

So here's something we're trying to do right now. I think we're working with Apple engineers, and my staff is also working aggressively to create some very unique widgets. We're looking at-- a phone directory which would have presence awareness and instant messaging built into it. We're looking at school calendars which will show students' lesson plans at the start of the day. We're looking at being able to quickly look at attendance reports, be able to look at the books that are overdue, and also alert students as to the new books that are available in library.

And finally, the little bitty handy helpers which go a long I think in schools we have some very large volumes of binders of information which deal with policies and procedures. We're able to quickly give our students, teachers, and parents, and administrators the policies at their fingertips, all with the push of a function key.

So in closing, I think I would like to call to action all the developers out here and ask their cooperation in building these true enterprise widgets. I think you can build widgets that not only extend to the educational enterprise, but also go and help other industries. For example, being able to get a dashboard view of all your key reports and statistics. Being able to look at your NOC, Network Operations Center, or other system management monitors. being able to look at your KPIs, or being able to quickly monitor your hosted applications, looking at bandwidth, and looking at effective utilization and availability. And finally, with presence awareness, and also looking at hardware integration widgets, you're able to manage and monitor your whole enterprise. So in a nutshell, what we see here is we have achieved our goal of enhancing or increasing the utilization of technology all with Apple products. And I'm very happy to say that we are the leading technology school district, and we plan on being there forever. Thank you.

Well, thank you, Vijay. And I'm sure that any of you who are in Fortune 500 IT shops find a lot of this very familiar in that the message of open standards really rings true here. Next, I'd like to introduce Joe Duemel from Aozora Bank. And he's the joint general manager there. And this is a bank. This is a more typical IT enterprise shop. And he would like to share with you some of his experience with Apple and with Mac and ideas about open source and open standards. So Joe, welcome. JOE MADRIGAL: Thanks, Bud.

So I'm sure the first question you have, since you're probably not up to date on all your Japanese banks, is what is Azora Bank? Azora Bank is a successful and profitable bank in downtown Tokyo. They're primarily a commercial and special finance bank. Azora recently turned over its management team. And as part of that, we looked at the challenge of how do we manage systems, technology, and people that were relatively stuck in the 1990s. Like some of the other presenters, I also have everything in my data center. I have stacks of IBMs. I have the old tape drives you see on old sci-fi movies. I have a pretty scary data center. The room that you're looking at there is a tape library. It's a robotic tape library. It manages a whole 1.2 terabyte of data. And it... Got a little robot in there, zips back and forth. It was replaced by something that looks about that high.

We have a new management team. Part of the charter of the new management team is they wanted to create an environment of knowledge workers. They wanted to get rid of a paper-based environment. They wanted to be an energetic or genki, is the Japanese word for it, bank. So they brought us on board and we looked at this thing and we said, okay, we don't want to deploy Windows. So we started with some rules. And the first rule is it's got to contribute to good karma. And nobody gets good karma with Windows.

We also discovered that we were really building from the ground up. As we started this process, we thought, okay, you know, we'll replace the desktops, put, you know, maybe some Linux in the back end, and we'll be fine. We started looking at replacing the desktops and discovered we needed to upgrade our network. Okay, I found out I had hubs underneath the floors. I wouldn't have put them there, but that's okay. We pull up the floors. We look at the floors. We're like, okay, we need a full recable. We start the recable effort.

We say, okay, now I need new switches. It's time to go to Cisco, get all new switches, do this right, because I don't want to do it again. As we did that, we looked at our power requirements. I didn't have enough power per floor to run my Cisco switches. So I said, okay, we went to our power distribution and we upgraded that. And they said, well, you only got about five, ten minutes of UPS. Okay, that doesn't really work either. We started pouring concrete in the basement to put in new UPSs so that we could put in the power distribution units, so we could put in the switches, so we could build a data center.

Fundamentally, what we were looking at doing is creating an IT revolution. We were going to replace every single desktop unit. We were going to take the 75 different vendors in our data center, consolidate them down to maybe four. We were going to streamline our systems and processes, make it a manageable thing, get rid of access databases and Excel spreadsheets, holding customer information, et cetera. So as we sat down and discussed it, we decided we needed some rules. The first rule was education. We did not want to end up being beholden to any vendor. We wanted to own our own environment. All our employees, all the guys who run the stuff daily are Japanese. You know, when you deal with a U.S. vendor, you have to make sure that you have the right education. So we made education a very high priority. We wanted our people to run their systems. Okay.

The second was design. I'll show you a picture in a couple seconds of what our employees work in, what their daily environment looks like. It's not so good. Design was really important to us. It's design of the work environment, design of the systems environment, design of the business processes. Design was important to get right because without it, we were back to the bad karma thing and it wasn't looking good. And the last was partnership. We knew that we couldn't do this alone. We knew that we had to go to the industry leaders, people like Apple, people like Cisco, people like EMC Documentum, people like TIBCO for integration. We needed this stuff to work and we needed to go to the people who knew it best. So as you look at this, this is a pretty big IT investment. We're spending $30, $40 million, tried to get this right, and that's just the first cut. So we wanted to do it right. We knew that partnership was gonna be really important. More rules.

First thing about partners, they have to be an innovation leader. They have to own their space. When you mention their name, people have to go, yeah, networking, they got it. Apple, yeah, desktops, they got it. They have to train our people in their intellectual property. Again, one of the things that really broke earlier was that it was really hard to separate our people from the vendor armies that kept giving me consulting bills. So it was important for me to actually really own my enterprise. Mutual nondisclosure. As you work in any of these partnerships, you find that you need to plan ahead five to ten years. Planning ahead five to ten years means you need to know what your vendors are going to do, and they need to know what you're going to do. So mutual non-disclosure became really important to us. And the last one is executive-level sponsorship. As you do a partnership like this, whether it's a partnership with Apple, partnership with Cisco, stuff's going to go wrong. It's not all perfect. It never is. It's important to be able to go to someone and say, hey look, help me make this work, and share in both our successes and our problems.

So as I described this to our management committee and as we worked together, the question we kept hearing was, "Why Apple?" The answer we thought was really clear. Undeniably they're the leader in the field in innovation, both hardware, software, both for the enterprise and the desktop. They really had it. The second was access to the best and brightest minds. It was really important to us to be able to go to the people at Apple that make things work and go, how do I do this here? Because it doesn't make sense. One of the things we looked at is as we were trying to pull all our systems together is, you know, okay, I got Java, I got Unix, I got 3270, I've got some legacy Windows stuff that runs in terminal servers, I've got X Windows, how do I make it work? The nice thing with OS X, it all just works. Don't have to do a whole lot different to make it work.

Highly internationalized. My kanji's really bad. I'm really happy that I can flip into an English screen. A lot of the Japanese employees are really happy that I'm not trapping them in an English screen. The fact that OS X is internationalized from the ground up, that Apple is built with a view towards being internationalized, is critically important to us. And Apple's really energetic, and that was real important.

So I told you earlier I was gonna show you what our employees work at. Basically, everybody has their very own 3 1/2 feet of linear space, which they share with whatever books and coffee cups they want to put on their desk. They sit next to each other. They're separated by a file cabinet. This is not a knowledge worker environment, so we destroyed it. Right now I have 20 floors in my headquarters, I have six floors in my data center, and I'm progressively destroying it floor by floor and replacing it. Here's what we're replacing it with. This is prior to our first rollout.

One of the scarier sites I've ever seen in my life, every one of those ended up VESA-mounted in cubicles. And as you VESA-mount them, you take the old stand off. We stacked the stands from one side of the floor to the other. It was kind of interesting. So some of the challenges we're looking at. We're a bank. We are legally responsible for our customers' information. It's very important in Japan that customer information never get out. We don't quite get the same option that they might have in the U.S. as saying, "Oops, I lost it." You say, "Oops, I lost it." They say, "You lost your banking license." My management says, "That's bad."

We have a high need of integration. I mean, we do have a bunch of systems. When we first looked at this, we were going to use Mac on the desktop, not in the server room. We had to get this stuff working. I have tons of paper on every floor. We're actually lightening our building's earthquake load because we're moving the paper into Documentum, which gets backed up into big, you know, terabytes and terabytes of storage, et cetera. But I needed to make Documentum, TIBCO, Oracle, everything work right. When we started this process, it didn't all work on Apple. So we had to go to TIBCO and say, hey, we need some help. TIBCO responded. They worked with Apple. Apple and TIBCO got together.

They got together a beta version of their messaging product, their system management tool. And I'm pretty happy to say that they're now generally available. They've been released, which means I can run them in production, which is a good thing. Also, TIBCO's also committed to porting the rest of their product stack to OS X by the end of the year. So that really helps me as well.

So the question I get asked from the Apple people is, "When are you going to Tiger?" So I'm kind of weird. My management asked me why I'm going to Apple. Apple asked me, "When am I going to Tiger?" The answer is I started out on Tiger. I had no choice. Tiger has the only security model that would work. All my clients are net boot with encrypted swap. I do not have any customer data on any disk on the floor. To me, that was absolutely critical.

It's that banking license thing. They seem to care. The second thing is card services. My security model is pretty tight. It's a combination of what you have, which is your physical key card, and what you know, which is your password. So basically, that key card also controls the ability to come in and out of the building. If you go to the restroom, you pull out the key card. The Mac flips using, I think at the moment, it's a user switch to a screensaver. You go to the bathroom, come back in, plug it back in, you're back in. The next thing is the ACLs. You've kind of heard a lot from them. They were critical for us. I do full role-based authentication. I know when anybody touches any data, and I really needed the ACLs and client lockdown based on roles.

The other thing that really, really hit us is manageability. Tokyo real estate's a little expensive. It was important for me to be able to put my help desk out in Fuchu. Fuchu's our data center, about 30 kilometers, maybe 40 kilometers outside of Tokyo. But I had executives, and executives need help. Um... It was important that my help desk guys could help my executives. So what we've done is every single map has got an iSite on top of it. We support video conferencing for every employee. We're kinda hoping to tweak the interaction model a little bit. And our help desk will use that. Our help desk basically will initiate a video chat, ask permission to take over the desktop with Apple Remote Desktop. And then basically show them how to start Word, start Mail, find their document for them.

The other thing that was kind of helpful, I mean, I've never personally figured out, you know, how do you manage 2,000 of these things? Apple has. Work Group Manager and Server Administrator or Server Admin have been great for it. One of the weird things that happened after we'd made the decision is some of our quants, and for those people not in finance, these are the people that really like numbers and are relatively boring to talk to. They came to us and said, we need you to build a supercomputer. And we're like, okay. Never quite done that. But instead, what we're going to probably do is use XGrid. And we're going to use XGrid and use some of the unused CPU models and then look at building a supercomputer. I don't think we can do as cool with video, but hopefully we'll get by a little bit. So here's what I wanted to end with. This is floor one done. This is the 16th floor in our headquarters.

I only have 19 more floors to do and a whole data center, but we're creating switchers 110 at a time every three weeks. So thank you very much. Thank you. Thanks a lot, Joe. So I think I like the switcher aspect of that. Next from RoTeC Healthcare we have Albert Prast who is both the Chief Information Officer and the Chief Technology Officer. And he's going to tell us about their experience. You may not have heard of RoTeC, but he'll explain that there is probably a RoTeC near you somewhere. And he'll talk about his experience with the Mac and with open standards and open source. So Al?

Thank you. Thank you. I'm glad to be here. This is the first time I've been to an Apple event, and I'm very impressed. by the attendees you you guys are great i've been to a lot of these things with other companies in the past and uh... this is fantastic so who is road tech and why haven't you heard of us we've been around for almost twenty years uh... were publicly traded company and uh... we provide durable medical equipment uh... services and medications uh... typically to homebound uh... patients The reason you haven't heard of us is our business model was started really providing service to medically underserviced areas of the country.

So, when I started at Rotec, I was told day one, really our motto is we're everywhere you've never been. Belzona, Mississippi, Coeur d'Alene, Idaho, Wheatland, Wyoming, some real places that are, we're in the middle of nowhere. Now since we've expanded, we have around 500 locations around the country right now and 5,000 employees. We kept the names of all of the companies that we acquired. We were an acquisitions company initially and we found out that a lot of people had a local presence and the mom and pops were doing a good job. So we have Hooks Healthcare and Revco and Quest and several others.

may have heard of some of those okay so that's what RoTeC does what do I do on a daily basis and there's a quote up there I'm gonna let you read it I'm like but I'm going to talk to it I go to a lot of meetings and I meet typically with end-users of the business I meet with consultants I meet with other folks on the senior management team and I meet with people on the board typically and I I walk in with charts very similar to what you saw with Vijay when he was up here, and I say, "This is what we're doing. This is what our landscape looks like." They always look at me and say, "You know, this is great, but you're making it too complicated. You need to simplify it." My answer is usually, "Okay, that's what I'll go do." Then I leave. I go back and I get with my guys and we go into a room and we say, "We need to simplify what we're doing." and they look and they nod and they say, okay, what do you want to do differently?

What boxes on the chart do you want to remove? And while we're talking about that, we usually have several emails come in, phone mails, project managers out working with different areas of the business, and we have middle managers and consultants running around with our customers in the field and coming up with new ideas. So we get constant requests, as most enterprise do, that moves into that extract, transform, and load area where we continue to have to provide more and more data and more and more ways of looking at the data. And I'm sitting in the room thinking we need to simplify, and at the same time I'm saying I've got a data warehouse that continues to grow and it grows and it grows, and the data sometimes has a tendency to get stale. That worries me a little bit. At the same time, we have many derivative businesses popping up that I didn't anticipate.

And I'm not saying didn't anticipate four years ago. Didn't anticipate a year and a half ago. Our patient base, who are our customers, are typically older folks. And we thought they liked the traditional model where they call in by the phone and deal with us that way. And I'm finding today I'm doing several hundred, close to 1,000 orders actually a day over the Internet. I didn't think I was going to have to do that. Telephony capability, moving faster than I could describe. A year ago, we didn't really do a lot with telephones outside of the traditional. I think probably many of you have heard of the McDonald's in Mississippi that when you go to the drive-thru, the voice comes on and says, Hi, what can I get you? What would you like to order?

And they are going to a data center someplace in Utah, a call center, who takes the call and by the time they get to the end of the drive-thru, their lunch or dinner is ready. We do a very similar thing today with taking orders where you'll dial a local location of ours and the call will be routed to one of our call centers in Murray, Kentucky right now. You'll speak to somebody there, they'll take your order, and by the time they finish, the delivery ticket to get your product to you will be printed out in the location. and or possibly drop ship if it's required. Again, we're still sitting in this room figuring out how to simplify. I also have this boat anchor, effectively I carry around with me all the time, which is my network. I have a 500-node network that I have to manage on a daily basis. Legacy applications in the teens. Scalability that, because of the changing business requirements, we don't really know from really quarter to quarter if we need to increase, you know, performance and capacity by 5% or 20%. We have cost pressures, just like everybody does. We have to be cost effective. And the uptime is, again, not a hand wave and something that I did want to mention. People a year ago, if we had an issue or a problem or some kind of malfunction with something as simple as email, that was okay. Today, users are angry if they get a piece of spam. And I mean literally really angry where they will call and they're very upset. It's hard to accommodate that.

So what are we doing and where are we going? I've tried to change our charts a little bit in the simplification when we left the room to say, I want our charts to look more like this and be more organized to process and specific type areas of the business, if you will. And we're one of the folks who have really adopted this SOA that you've heard about, the service-oriented architecture.

We have deployed an ESB, which we did probably third quarter last year. and we use every day all of the acronyms that you've come to. Many of you write. A lot of you have heard about. Some of you are just looking at it. But the WSDLs, the UDDI, the web services, XML, XPath, XQuery, those are things that my folks are very, very familiar with. We released our first application, if you will, a rich internet application built entirely upon this architecture at the beginning of this year. And what you see on this chart represents there's four legacy systems of our partners that are represented on this chart, two external web service providers, four internal systems, three different database servers behind our firewall, and one rich internet client. The application that we rolled out was called e-intake, and it was one of the best received applications we've ever rolled out. And I've had users cry before when we've given them things, but I've never had them cry with tears of joy. And honest to God, I really had a end user in a billing center come up to me and hug me. She had tears in her eyes and say, "You have no idea how you've changed my life." And I wasn't expecting that, but it's better than getting shot at, so that's a good thing. Okay, so how are we doing it, and what are the technologies, and where are we looking to go, and what would we like you guys to hear from us? We have lots of different databases, and we've just put a couple up here, and there's a lot more, and we know that, and we've got these application servers, and we know there's a lot more on top of that. But effectively, the message I want to give you with this slide is, Everything that we're doing now is layers of abstraction, okay? So we're using everything as a building block. So we have these databases available, and we have tools that we can put on top of the databases to make them look to one of the developers really the same, even though they may be different types of databases. Now, they each have their own great features and functions, and we take advantage of those where we can and when necessary. On the application server side, we have a lot of different products that we could use there. The great thing is we have code that we can write one place and deploy really on anybody's application server. More importantly, we also can use very sophisticated types of development using something like a JDeveloper or WebLogic for some of the more deep business logic, and we could use lighter-weight tools if we'd like to to write code that runs on the same platform. So that's kind of what our environment looks like for what we're rolling out. Okay, so why am I here and what are we doing with Apple? So because we have these layers of abstraction and we can build on all these different things, the guys who develop and work in the data center finally come to me and I say, this is great. We have a new way of doing things. We have some new products that we've rolled out and now we need to roll it out on something. Who's the vendor? And I ask that question because I've been doing this for a while and everything we've done in the past, we've been usually given a vendor, "Oh, to do this, you have to work with so-and-so." So if you look at my data center, it's not unlike most other enterprise data centers today. There's a lot of black, there's a lot of gray, there's a little bit of silver, et cetera, et cetera.

We have a lot of different vendors in there today. But it was different with where we're going now because the answers really changed a little bit. You can use anything you want, almost. I can go with Microsoft Intel. I can go with Linux Intel. I can go with IBM AIX. I can go with Sun Solaris. And oh, by the way, now we can go with Apple and OS X. Mm-hmm, that's interesting.

Okay, so what should we do? We've never had that many choices before. How do we choose? So we start looking at a lot of different things, and obviously price performance is a big one. Scalability is a big one. This cluster computing capability is now coming on the scene, and that's great. So we took a look at really everything side by side, and then we kind of found that, you know, there was a short list, and Apple was certainly on it, and one or two of the other vendors. And then you ended up going with feel. And the analogy that I gave yesterday is you don't really get a rule book for how to be a parent for those of you who have kids. After a while, you just learn that when things come up, you learn what feels right, and you learn how to answer questions that you have confidence in and you say, "This isn't gonna keep me up at night." And that's one of the things that we realized with Tiger and these--the hardware, is that it just works, it scales. I can start with five nodes that we have right now for this application, and it performs very well. But I don't have to worry about it if the demands get, you know, greater and greater and greater because we can just keep adding machines to it. So that's great. So we've decided that we're gonna use these for the servers. Now, what do we do with disk? Same sort of thing.

Do I have any specific vendor that I have to use that's a requirement with this? Well, no, we can look at EMC and IBM and NSTOR and Apple, and again, from a price performance standpoint, this stuff's really, really good. Okay, so what's the downside? Is there a single point of failure? No, we can configure it so there's no single point of failure. Okay, you know, how big can it get?

It can get bigger than we need to get in the foreseeable future with storage capacity. So that was kind of exciting, and that's why we selected Apple for the storage part of it, and we're very happy with the solution today. Now, I want to close with something that's even more important, and you've heard from every single person, or at least every speaker that I've heard today, we hear about innovation. The fact that we now can choose any piece of hardware or any vendor that we want to to be able to run these things is great, and it gives us a lot of flexibility as enterprise users. So who's gonna win is the most innovative folks. So I'm here today because of that. I've never been here before, and I hope to see you all here next year to tell you about Chapter 2 of our story and where we go next with Apple. Thank you.

Thank you. Well, thanks, Al. And so I think you've seen some good examples from customers, opportunities that have been opened up for us with things like Oracle 10G server available and customers starting to deploy that. I think if we stand here next year, you're going to see even more of this.

Now, what I'd like to close out with is just point out a few things that you should absolutely know about OS X server on Tiger and its role in the enterprise. I'll point them out. There's a lot more you can learn at the sessions here, but I'll just call out a few. I've mentioned 64-bit computing. If you look around your shop, you are going to have applications that need 64-bit address spaces that have 16 gigabytes of physical memory attached to a single server.

That's supported in Tiger. ACLs, we follow the Microsoft model. So if you're in a mixed environment or, as you heard, you need fine-grained control over who has access to what, ACLs provides that with a full inheritance model with very fine-grained control with groups, et cetera, et cetera. link aggregation I mentioned a little bit this is a very cheap way to get redundant very high speed just getting up to gigabit Ethernet's you'll get two gigabits of throughput if one of them goes down and still keeps working on top of this plumbing we have a plethora of services I chat server is one of my favorite you can now run I chat inside your corporate firewall provide instant messaging between your employees. It's encrypted. It's authenticated with either passwords or Kerberos. So you can now leverage instant messaging as a productivity tool for your employees. Included, no extra charge with Tiger, based on Jabber, which is an open source instant messaging solution. So it's also compatible with all the other Jabber-based solutions out there on other platforms.

Software Update Server, I mentioned, puts you in the driver's seat for when people get updates showing up on their system, saves you bandwidth so that you're not downloading repetitively all of the updates that we put out onto every desktop or every server. You download it once to your proxy and then distribute from there into the enterprise. Weblog Server, blogs, as I mentioned, that's the new groupware for the next decade, deploy blog servers inside your enterprise. This is based on Blogsum, which is a Java-based blog server. Not only is it blogs for individuals, but blogs that can be targeted for groups or projects and run securely, again, behind the firewall. Or for people who want to blog out to the world and let their employees talk to the world, you can do that too.

Junk mail and virus filtering. Spam Assassin is the open source project that we have integrated in. It's a very high quality spam killer, both in terms of adaptively identifying spam and getting rid of it at the server, and taking blacklist feeds to remove spam from blacklists which are available on the internet. Virus protection via Clam AV, another open source project integrated in with our mail server. We'll return or delete or quarantine or whatever you want to do with those viruses if they come in. Not that those viruses will affect your Macs, but you don't want to be passing them on to Windows machines as well.

Xgrid, you heard an example of, you know, IT, CIO gets told, hey, we want a supercomputer. Our quants need something like that. Take a look at Xgrid. It's built into both client and server, so it's there on every single system of Tiger, whether it's server or client. Secure, you don't have to run a node as an Xgrid node. You can enable it, though, and it gets authenticated and will only run the jobs that you want it to run. And then finally, XSAN version 1.1, available on Tiger. So it's really XSAN for Tiger. It's available for Tiger and Panther. The benefits are large-scale storage consolidation. So 64-bit file system, up to 2 petabytes in size of storage, virtually no performance or volume or scalability limitations in this, literally hundreds of XSERV RAIDs for increased performance. And best of all, it's a free update. All users of XAML 1.0 will get XAML 1.1 for free.

And to wrap up, I just want to call your attention to a few very good sessions to go to. Apple Data Center, First Flow Moscone West, right down below. Take a visit there. Look at an example of an Apple IT data center. Interactive panel discussion at 7:00 p.m. on Tuesday. Don't miss that. O'Reilly brown bags are great. If you've got an interest in Python, PHP, Perl, MySQL, I love O'Reilly. O'Reilly's great.

Go to the brown bags. Late night labs. Again, Apple experts in areas of Mac OS X server. How do you put a Mac OS X server in a heterogeneous environment? What's Apple Remote Desktop all about? How do you deploy in XAN, web objects, et cetera? So these late night labs are incredibly valuable time you can spend with Apple experts in these areas.

And then finally, Oracle Database 10G on Mac OS X server Friday, June 10 at 10:30 at Second Level, Alcove 3, across from Enterprise IT Lab. Talk to Oracle people. Learn about Oracle 10G on Mac OS X. And finally, as I mentioned, 40% of the attendees are IT related at this conference. And so we have a ton of sessions. They're in your notebook that you got handed out with your badge. So I'm not going to go through all these, but I'll just click through them. And finally, thank you very much for coming to Worldwide Developer Conference 2005.