05282008 Bernstein Strategic Decisions Conference Ray Ozzie by leader6


									Who: Ray Ozzie, chief software architect, Microsoft

When: Wednesday, May 28, 2008

Where: Bernstein Strategic Decisions Conference - New York, New York

CHARLES DI BONA (Host): Thank you all for coming. I think there's still a couple
open seats in the back perhaps, but we're really happy today to have Ray Ozzie here
with us, who is the Chief Software Architect at Microsoft. And it's sort of traditional
that we do a quick bio of people as an introduction here, but if I were to go through
Ray's whole software history, we'd be here through lunch. So let me just hit a
couple of the highlights.

His software record goes back even to Data General where he worked on some
distributed operating systems, and on VisiCalc later on, and Lotus Symphony. And
then as the founder and president of Iris Associates, he led the development of Lotus
Notes. And then subsequently, in 1997, founded a company called Groove which
worked on sort of next generation collaboration software, and that company was
subsequently acquired by Microsoft in 2005 when Ray joined Microsoft as the Chief
Technology Officer.

Since June of '06, Ray has been the Chief Software Architect. That was the time
when Bill Gates announced that he would, in two years, at the end of this June, be
stepping out of an operating role at Microsoft and focusing his attention more on his
foundation. So we're about a month away from that. And according to the company
Web site, Ray's responsibilities "include oversight of the company's technical strategy
and product architecture," which is an extremely insightful job description.

So maybe that's a good place to start, Ray, maybe in your words could you tell us a
little bit of the framework here for how you view your job, sort of the kind of things
you think about on a daily basis, and how you interact with the operating executives
at Microsoft, and with Bill, and with Craig Mundie and others.

RAY OZZIE: Sure. Well, Microsoft is a big company. We have a lot of products.
Microsoft is really broken down into three divisions, kind of one that has Windows
and Online Services, one that has our productivity activities, and one is
entertainment and devices. And these groups concentrate on growing those
businesses. They understand their markets, they understand exactly what next
steps to take with their products. The role of Chief Software Architect is kind of an
interesting role relative to the divisions, because in essence it's one where one can
step back and look externally at trends, and do pattern matching as to what's
happening across the entire environment. This is kind of an important role in our
industry because our industry moves quickly, and it moves progressively over a long
period of time.

I've been in the industry, as you've said, for quite a while, and I've been fortunate
enough to have been through a number of fairly significant transitions because of the
steady progress of technology from mainframe to mini, mini to PC, PC to graphical
user interface PC or LAN, and then the Web, and now we're undergoing a number of
transformations, a transformation toward cloud computing, a transformation toward
mobile devices. There are just a number of things that are happening.
This transition toward cloud computing and toward services is really what I've been
spending the vast, vast majority of my time on since I joined Microsoft and that has
only really stepped up since I assumed the CSA role. When I came onboard, much
of the company, many, many of the developers of the company were working on
what would eventually become Vista, what would eventually become Office 2007,
and I began by kind of writing and communicating to various groups in terms of the
macro level trends that I saw happening. As time went on, as those products
shipped, I've spent a lot more time much more hands-on, much more directly with
the product teams trying to map out what specific scenarios are important, how
those product groups' products would fit in with one another in this transformation
toward services.

The interrelationship with myself and the division presidents is kind of very
interesting. We're more or less partners at the senior leadership team level. Craig
Mundie and I, Craig and I both split Bill's role. I have taken on more of the here and
now, Craig has taken on research and policy, which is more longer term. He's got
more of an external role, I'm more kind of the man behind the curtains so to speak,
but I spend a lot more time hands-on with the product groups.

CHARLES DI BONA: Just so I don't forget, there were on your seats question
forms, so if you have questions, please write them on those, and I guess pass them
to the -- to the outside aisle or the center aisle -- the center aisle, and we'll sort of
work them into the presentation as they come up. So feel free at any time to do

I guess what you've been talking about here, there's a lot of speculation about cloud
computing, and about software as a service, and about computation moving onto the
Web. And clearly Microsoft has articulated something of a different vision about
software and services, where you blend sort of the more traditional forms of
computation with the new service-based offerings.

Before going to the Microsoft vision, maybe it's helpful to sort of circle back to cloud
computing, and software as a service, those are terms that get thrown around a lot,
and in very loose ways, and I think that they've taken on multiple meanings, and
maybe it's helpful for you to sort of say how you think about those trends and what
they really mean?

RAY OZZIE: Sure. Well, it really comes back to understanding what's going on with
the technology and projecting out and kind of pattern matching what's happening in
different segments of the market, or what might happen based on those technology
trends. What's going on right now is really based on the confluence of cheap
storage, cheap computing, and probably more than anything cheap, ubiquitous
communications amongst businesses, or homes and businesses, and data centers,
and so on. And the confluence of all of these things gives us, as architects, the
opportunity to kind of figure out exactly where is the best place to put computation
and storage to solve a given problem for specific business customers or for
consumers. When I'm working with different product groups, how these things
impact their customer bases really may vary, so I've really come up with three
principles that I use to talk to the groups that help them understand and help us kind
of have a great dialogue about how these things might impact specific solutions.

I've come up with kind of three principles, the first of which is that the Web is really
the hub of a device mesh, and a mesh of people. And what I mean by that is,
formerly or traditionally when we bought a PC, we bought it as a standalone device.
When we buy a phone that phone is connected to the network, but it's kind of
independent of other phones and the PC. When we would buy a TV that TV is
separate from the various other things that we have. In essence, now that we have
the Internet as kind of the global thing that can connect us all together, and now
that many, many of these devices are coming with wireless networking built-in, with
Ethernet jacks built-in, it gives us as architects the ability to actually bring them all
together. And many of us have challenges configuring these various devices, many
of us have multiple PCs, where there's the PC at work and a PC at home, a couple of
PCs at home. You try to get media from one to another, you might want to get
applications from one to another, or data. The Web, the fact that we have global
connectivity, it affords us the ability to kind of think of things a little bit differently
because we can use the Web as a hub to bring it all together.

The other thing is that it wasn't too long ago that we really didn't have a way,
electronically, for all of us to communicate with one another, and to connect, and to
rendezvous with one another. I spent a large part of my career working on
collaboration software, and one of the most fundamental elements is getting people
within a company or people between companies to talk with one another. It used to
be extremely challenging. Now we all take for granted the fact that if you just have
somebody's e-mail address, you can send them something, you can get something.
And just by reassessing offerings and saying, if we have the Web as the one place to
bring all of our devices together, and all our people, all the people who use them
together, it enables you to reframe the solutions that you're trying to do across a
number of different markets.

The second principle is really something that I refer to as the power of choice as
business is transitioning from using just data centers to being able to use cloud-
based computing. Many, many years ago, when I started programming back in '68,
I used a time sharing system, GE 400 computer on some time sharing system in
Chicago, and it was a 300 baud line, you know, with a teletype going to that
computer. And you can't do a whole lot with a 300 baud line. And so companies
have built up over the years data centers so you could have co-location of the people
who were using the solutions and the computers themselves. But now with this
ubiquitous bandwidth that we've got where we have these really, really high speed,
low latency networks amongst businesses, and between businesses and potentially
large scale shared data centers, we can now choose what do we put up there in
those data centers versus what do we keep onsite. And enterprises now have the
power of choice. They can decide what are the strategic systems that they really
need control of, what are the things that are more infrastructural in nature that they
might be able to get some economy of scale by treating them more tactically and
offloading them to someone else to manage. That's a very big transformation that
impacts our company and our industry and our customers.

The third real trend is something that I refer to, it's for developers, and I refer to it
as small pieces loosely joined, and what I really mean by that is that traditionally
developers have built programs as big, monolithic hunks of software, a big program
that you install on your PC, a big a program you install on a server. And the world
that we're moving into is fundamentally different. It has many, many computers,
little processors, on one machine. It has many, many computers up in the cloud in
order to get broad scale and robustness. And the tools and the techniques that
developers use to build software, to target a user using multiple devices, not just a
single device, are dramatically different. And that's new opportunities for a company
like Microsoft in order to serve developer tools and platforms to those developers.

So fundamentally, I just think that kind of stepping back to the role of CSA, one of
my jobs is to kind of look at these trends, and to say where are there threats
because of inflection points, competitive threats, or threats of disruption, and where
are there opportunities. And this is one where there just seems to be tremendous
growth opportunity from the user standpoint, from the enterprise standpoint, and
from the developer standpoint because of those transitions.

CHARLES DI BONA: And so maybe it would make sense to sort of now delve into
software and services, and how is Microsoft sort of addressing these opportunities,
and these threats, specifically how is your architecture differentiated from
competitors, obviously Google springs to mind? And sort of what is the value,
ultimately, to the IT department, I guess, would be a large buyer, but also to the
consumer, and to the other users, why would they buy it?

RAY OZZIE: Well, kind of going back to the principles, if you take these principles
of what's going on in terms of using the Web as a hub, and the power of choice, and
these architectural principles, you can see how it would net out to opportunity if you
kind of go on a market by market segment, because we are a broad company, we're
very diversified, we have different products for different markets.

On the consumer side, you know, it really is powerful in a number of ways. Actually,
let me just back up, from the standpoint of an OS, you know, Windows and Office
are kind of the franchises that people are most familiar with of Microsoft, and the
very nature of an OS really changes in the world moving forward. We've
conceptualized everybody who has used an operating system on a personal
computer, or IT people who have used operating systems on servers think of it as a
piece of software that operates one PC and one computer.

But the opportunity moving forward is for us to kind of step back for a moment and
conceptualize, if we were building an operating system today, what would it look
like? It wouldn't really look like just one computer. We all deal with many, many
devices in our lives. As I said before, I deal with multiple PCs at home, my phone,
my TV, my car has this navigation system, we have media players, and the
opportunity is to really ask the question, what would an OS look like in a world of
multiple devices, in a world where instead of the computer being at the center, you
are at the center, and how do you bring these things together to have them all do
what you'd like, to have your e-mail and contacts, or your media, or your photos
synchronized amongst them so that you can use the device that's appropriate in a
given situation. And each one of us will choose some different mix based on the
nature of our lives, our preferences, whether we live in a culture that's a commuter-
based culture, where we might use mobile devices differently. If you spend a lot of
time in mass transit versus driving, the types of devices you might use are different.
So I think the first opportunity is really at the redefinition of the OS.

The second would be in the redefinition of productivity into connected productivity.
We've classically thought of productivity tools such as Office as being tools that
operate on one device, one PC, and it is an amazing, amazing piece of software that
lets you create and edit, and present documents.
PCs are becoming more powerful, they have bigger, and bigger screens. But, the
Internet also offers new functions that are complementary to what's been going on in
PCs. The Web is very, very good at sharing, it's good at universal access and
sharing, it's not necessarily as good at presentation and interactivity. So it
complements the PC in many ways. So the opportunity for us, in the realm of
productivity, is to redefine the scenarios, to step back and say, how can we grow
Office, in terms of what it really is, from both a creation and editing perspective, and
the sharing scenarios.

In terms of mobile devices, also, again, using the Web as a hub, the opportunity is
really not just spreadsheets, and documents, and presentations on phones. Phones
have cameras built in, phones have GPS built in, some of them have little ways of
doing note taking, whether it's through inking, or through a little keyboard. We have
them with us all the time. Those phones can be woven into higher-level scenarios
involving note taking, so that when you take the picture of that whiteboard, it
automatically gets synchronized to the PC up through the Web. So, again, in the
productivity realm I think there's a growth opportunity in reaching new types of
usage scenarios, broader markets, new form factors of devices, and so on.

In the entertainment realm, we all use digital cameras, digital camcorders, we have
televisions, some of these televisions are now IP-enabled TVs, gaming consoles. The
opportunity in entertainment and media is really, again, to weave these things
together into higher-level scenarios that bring all these devices together. And I
already talked about the opportunity in the enterprise, in terms of connected
business, and so on. So again, I think the broader opportunity is essentially to
understand how these different markets grow based on the conceptualization of
more than one device, and more than advanced scenarios than just the more limited
way that we've thought about them in the past.

CHARLES DI BONA: So if you think about the sort of competitive environment as
we move into this more connected world, who do you see as your big competitors
going into this space, and what kind of assets do they bring that maybe you can't
replicate, or don't have replicated yet, and how do you deal with those issues?

RAY OZZIE: Well, everyone starts from where their core strength is, but I think the
main thing that we should all realize is that this is an industry-wide trend, and we
are all going to end up in the same place. In any given inflection point, industry
inflection point, whether it was mainframe to mini, mini to PC, and so on, the new
guy comes in and he tries to disrupt the incumbent by saying the world is going to
change, it's going to dramatically shift to this completely new model, and there's
going to be no middle ground, it's just going to go all the new way.

In today's world those are the people who say: You're going to dump your servers in
your enterprise, and move it all to the cloud; you're going to no longer use a PC,
you're only going to use a browser; you're not going to use a phone anymore, you're
only going to use a browser. And the incumbents who have their head in the sand,
and think that a transformation is not occurring, will be disrupted by those folks, but
when all is said and done, in each one of these transitions the aggregate market
always ends up larger than it was before, because people find new uses of the new
technology, and the old ends up getting adapted into the new.

We believe, fundamentally, in a hybrid model. And that's why when you hear us talk
bout services, you hear us say software plus services, because we believe that the
real opportunities moving forward are PC plus phone, PC plus mobile Internet device,
Web plus PC, as opposed to just everything in a browser. I think we have some
unique core strengths in that we have both the enterprise business, and the
consumer business under one roof. The consumer business in MSN has given us the
unique skill of understanding how to do very, very high scale services, services that
serve hundreds of millions of people, whereas as a an enterprise-only player we
might not have been able to recognize the fundamental rearchitecture that's
necessary to do that kind of scaling.

The flip side is, if you're only serving consumers with these high-scale things, you
don't really realize how hard it is to be an enterprise player. We started as a
desktop, in the desktop business, and it took us many, many years to gain
enterprise credibility, and really to understand enterprise requirements, and
understand solution selling, and understand really what an enterprise needs in terms
of touch. I think the advantage that we have is that we can bring those things

CHARLES DI BONA: I'll sort of -- the marquee, or the poster child for the Web-
centric view is Google, clearly, at least in people's mind. Are they more difficult,
because of their capitalization, because of their sort of intellectual capital, to
compete with than your sort of historic competitors at Microsoft, and also do you see
them sort of with Gears, and with the Mini, and with Android, sort of reaching out to
try to get onto the client side, to get into those other -- onto the devices, and into
the other parts of the ecosystem that they're not in yet?

RAY OZZIE: Sure. Well, as I said, I think the software plus services is an industry
trend, it's not just a Microsoft trend. They're introducing client-side things, just like
Salesforce has an offline client edition, because that's what customers want.
Customers who want to deploy these things have individual users with laptops, those
users like their laptops, they like to carry them around, they like to have data locally,
they also like to have centralized services, and all the advantage that services afford.
It's a world of mobility, so everyone has got some kind of a play in mobility.

I think Google, they're a tremendously strong competitor, very, very capable, very,
very smart people. The fact that they generate the amount of cash that they have
leaves it up to leadership, and how intentional they are about where they choose to
compete, and where they don't, and from that perspective they could compete in any
market that we're in, and from that perspective, of course we take them very, very
seriously. But, Microsoft is a very interesting company in that it's had to, by
necessity, kind of build up a culture of -- I guess a culture of crisis. Ever since the
early, early days Microsoft has always faced some amazing competitor that looked
like it was going to be some roadblock to success.

When I was at Lotus, we were a bigger company than Microsoft, and it was hard for
anyone to imagine how Microsoft could end up getting credibility in word processing,
in spreadsheets. We were very daunting competitors, we had so much more
progress. We had spreadsheets on the mainframe, 1-2-3-M, we had spreadsheets
on the VAX. It takes perseverance, it takes investment, it takes an understanding
that you can't compete by just chasing tail lights, you have to find some way of
flanking, or appropriately leapfrogging, because when you're in the face of somebody
who is running very fast and they're ahead, you have to find alternative ways.
Each one of these big competitive battles over the years, against Sony with
Playstation, open source, has left the company more resilient, and more well
positioned than before that competitive battle, and the outcomes are more diverse.
If you look at the competition with Playstation, it really formed the basis for our
entire connected entertainment strategy. The productivity wars in those days has
evolved and now we've got amazing opportunities in voice over IP, and other types
of collaborative infrastructure that couldn't have even been envisioned at the time.

Open source, that competitive battle, which was just -- in many ways it was
potentially much more disruptive than Google, who has to make a profit, and has to
report earnings to shareholders, open source has made Microsoft a much stronger
company, in terms of really understanding the importance of interoperability, and
serving our enterprise customers who have heterogeneous systems within their data
centers. So, yes, Google is a very strong competitor. Yes, we've got a lot of work to
do, but the company really is used to, culturally used to dealing with these kinds of

CHARLES DI BONA: You talked about sort of not following the headlights,
leapfrogging in these situations, clearly a lot of the questions already coming up here
are about Yahoo, and about how do you establish yourself in probably the weakest
part of your business, relative to that big competitor. What does the post-Yahoo
world look like, or is it even a post-Yahoo world, frankly, at this point? How do you,
from your perspective, see Microsoft's capabilities to sort of extend that franchise if
there isn't an acquisition of a lot of traffic?

RAY OZZIE: Well, just to put things in perspective, before I directly answer the
question, people don't like to talk out it a lot, but Microsoft has a very, very strong
position in the online space in some very important dimensions. We have 475
million people who use Live IDs, which are the basis for all of our online properties.
We are number one in terms of minutes, in terms of engagement of users in our
communication properties, even beyond the online properties. If you can just think
of the level of engagement that people have with things like Outlook, and Office, in
the other parts of our business, user engagement is a very, very core important part
of anything that you do in the online space.

I think it's really important to understand that there are three underpinnings of being
an online company. Number one is, user engagement, because essentially those are
the first party properties that represent potential places that advertisers could
advertise their products and services. The second thing is search, and our strategy
is to innovate and change the game in search. Search is a very, very important first
party property, and it's important to advertisers just like other highly engaging
properties. And the third leg in the stool is really an ad platform, an ad platform
that's of high scale, and an ad platform that presents a real competitive choice out
there in the market.

So user engagement, search, and a very strong ad platform are all essential to a
powerful, thriving online business. And we are innovating in search. We've got
some of our best people at Microsoft working on search. We're behind. There's lots
of opportunity, but there's lots of opportunity to innovate.

I know it's kind of hard to think of being in the early days in search because we've all
got our habits now, we go to our search engine, we type the words into the box
based on what we're looking for, but when you're trying to arrange a vacation,
potentially arrange it with other people. When you're looking up some information
related to health because of a friend or a relative who might have a health issue, I
think any one of us, just from our personal experiences, could look at it and go wow,
there is room for improvement. There's a world of information out there, and how
we structure our request in order to find the information that we want, there is
plenty of room for improvement and innovation. And I think we really haven't seen
any high scale implementations of really good innovative interfaces that have come
out yet, and we are investing in those areas. With Cash Back you've seen that there
is room for innovation in business model, and I think there are many, many different
ideas that different people will have over the next few years in terms of innovation in
business models that will help the end user, that will help the advertiser.

In terms of the properties for engagement, there's tremendous, tremendous room
for innovation. In the early Internet era, content, commerce and community were
really the three legs under the stool of user engagement. And communications like
instant messaging and e-mail was what transpired in the early days. Content,
portals such as Yahoo, and MSN, and AOL, were the simple way that content was
presented in those early days. Now content is being mixed in with communications,
and if you look at social networks, the photos that are shared on those Web sites,
that's the next generation of communications assets, and community is being
intermixed into content sites, and things like Digg are showing the fact that content
can be ranked based on measurements of community behavior. There's lots of room
for innovation in terms of user engagement.

So, in any case, Yahoo was not a strategy unto itself, Yahoo is an accelerator. We
view Yahoo as an accelerator to the ad platform, potentially to the user engagement,
and so on. We'd love to still discuss possibilities with Yahoo, but beyond that I don't
have anything to talk about.

CHARLES DI BONA: It sounds like, and there have been a couple of questions
about this as well, as you sort of move into a world, let's leave Yahoo out of it, as
you move forward into this, from your perspective, what is the value of partnering
for some of this versus building it in-house versus acquiring, how do you see that
sort of array of options particularly in the online space?

RAY OZZIE: I think it's extremely important. We are very, very serious about the
online space, and in order to have a very, very healthy ad platform, as I said, you
need a powerful search, and you need a large amount of user engagement. We
cannot, no single player can produce all the content that people are interested in.
The Internet brings us all together, but then we self-select into small segments
based on our own interests and demographics, and the Internet is a very broad and
interesting place, and it's going to get more and more interesting over time. And so
forever there will be many, many different publishers that have content and other
activities that users are interested in doing. In order to have a successful ad
platform, we need to make that ad platform appealing and profitable for those
publishers in addition to us as first party publishers. So we do need to build first
party content, we do need to have strong first party engagement as a part of
building that ecosystem, but we also need to partner very, very strongly in that

CHARLES DI BONA: There were a couple of questions about this as well, software
and services, I think you talked a lot about the opportunity for software plus services
to sort of drive incremental revenue, but there are some defensive characteristics.
In fact, I think some people would argue that it's primarily defensive. But maybe
you can talk about sort of the threats to the core business, and how software and
services sort of addresses it?

RAY OZZIE: Sure. Let's just say, we could focus on any one of those markets that
I talked about earlier, but just take the enterprise environment. We are a purveyor
of enterprise infrastructure, Exchange as an e-mail infrastructure, SharePoint as a
content management and applications server infrastructure, and so on. And in any
of these areas, we have very strong competitors, a very strong class of competitors,
Notes in the e-mail market, and so on. At any given moment in time, a customer of
our product will be looking at the environment, just like we are, and saying, should I
continue to invest in you, should I invest in your competitor, or what will be the
catalyzing event that will take me to evaluate some completely new architecture for
something like that?

Different people have different requirements. In financial services, for example,
there are requirements of compliance that might enter into the decision as to
whether to run your own infrastructure, or to host it elsewhere. Other companies
might say, hey, that's not my core skill, I'll just outsource that and use my well-
trained IT staff to work on business applications that are more strategic to my
business. So for some businesses, essentially, it's we want to be there from a
defensive perspective to make sure that when that customer who might want to
move things into the cloud wants to do so that they don't have to make some
dramatic new change that implies retraining of users, retraining of IT, we want to
give them choice to run Exchange in the cloud or Exchange on premises. So we
have offerings, Exchange Online, SharePoint Online, CRM Online, that let them retain
those skills, retain the user skills, and run those things in the cloud.

Similarly, there are some customers, such as Coca-Cola, who we've been trying to
win over for years as customers, they're Notes customers, and they decided that
they wanted to evaluate the cloud, and they wanted to use that as a point where
they just reevaluate everything, and they are now new customers of Microsoft's
Exchange Online. They're initially moving 20,000 users over of ultimately 75,000
total users. But they're using the fact that these online services are available as a
catalyst for that decision, and so that represent new opportunity, new growth
opportunity for us that arguably we might or might not have been able to capture

So it's a combination. I said earlier that at every one of these technology inflection
points, it doesn't just shift from one to the other. It ends up netting out to an
aggregate of growth, and there are both defensive and growth elements in the
services transformation.

CHARLES DI BONA: Maybe we could shift gears just a little bit, because one of the
other hot topics in software is virtualization, and maybe you could sort of talk about
that in the context of software plus services, and your strategy, and how you use
virtualization in your own data centers, how you see it playing out in corporate data
centers, and even sort of onto the client side as well?

RAY OZZIE: Sure. Virtualization is a great technology. It's been around since the
late '60s, early '70s. I think it was CPCMS, or there was some early products that
were the predecessors of VM-370. It's a very well proven technology. It's a
technology that enables isolation, the simulation of multiple machines on one
potentially larger machine. And much of the job of an operating system itself is to
let many things go on concurrently, this just takes it to another level. And ultimately
virtualization is really just a feature of the OS. You can think of the OS as either one
thing that's operating multiple tasks that share memory, or many things, but it's
really just juggling the use of one piece of hardware to do multiple tasks.

That technology has applicability both on the desktop and in the data center, and it's
got operability both in the enterprise data center and in the cloud. In the enterprise
data center, it's really mostly about making the best use of that capital investment
that you've got. You know, you want to run those applications as different
workloads. Some might have a heavy I/O mix, some of those workloads might be
very computational. By putting those together on one box, you can actually make
the best use of that box. And so from an economic perspective, it's very, very, very
important for the enterprise.

In the cloud, it's actually not just important for that reason, but there are other
reasons that really aren't written about very much yet, I think they will be overtime,
but we build these huge data centers out in the middle of nowhere, co-located with
dams, and in other places sources of cheap electricity, and we sent up racks of tens
of thousands of computers, and spend a lot of money and energy on cooling, and the
goal of that data center, our goal, is to make the best use of that, and ultimately
there's an environmental issue of if you're cooling these things, if you're spending all
this money on power, you want to make sure that every last cycle of those data
centers is being used very, very efficiently, and very, very effectively.

Microsoft today uses this to optimize the use of our hardware in the data center
across many, many, many of our properties, and over time, as you can see with our
customer environments like Exchange up in the data center, over time we're
extending the use of our facilities to more, and more, and more third party
customers and developers. So that technology is very important in that realm.

On the desktop, it's also very important, but more in the realm of compatibility. It's
really the ultimate way of assuring that if you have written an application for a
desktop environment, that you can make it run from one version to the next, to the
next of an OS. So I think what you'll see is as we make further and further
improvements and refinements to the desktop OS over time, we'll make use of
virtualization technology to ensure that compatibility. Some out here might be
familiar with an acquisition that we made some time back, Softricity up in Boston,
that's desktop virtualization software that lowers desktop management cost by
enabling an enterprise to package up an application and use that virtualization
technology to ensure that they can get it on a broad variety of desktops running
different OSes and different versions, and maintain absolute compatibility.

CHARLES DI BONA: I think you sort of touched on what I suspect is your answer
here, but there's been a lot made about disintermediating the OS with virtualization,
essentially running straight onto the virtuals on the hypervisor. Can you discuss sort
of the shortcomings, or your view of the shortcomings of that, or the potential for

RAY OZZIE: I don't think it really disintermediates the OS. Certainly, you can run a
hypervisor of one vendor and an OS from a different vendor. But, really -- or you
can run multiple from the same vendor. Ultimately it all comes down to what is the
best way to run the workloads that are relevant to a customer on a given physical
piece of hardware. So Windows and Microsoft's hypervisor will run all of the
Windows related workloads, and non-Windows related workloads that are relevant to
a customer on that piece of hardware.

CHARLES DI BONA: Going back to something you said earlier in the discussion
about operating systems sort of basically being redefined from being a device
specific, to a much more broad-based, running multiple devices, where do you see
the line when it starts to impinge on what the network does? Where do you see sort
of that line between sort of network management and operating system functionality
ultimately falling?

RAY OZZIE: There is kind of an unusual, kind of fractional nature to what's going
on, on a given chip right now, increasingly. We started out with one core, one CPU
on a chip, and now on most computers we've got two or four cores. We're heading
into a world right now where the clock speed of a given CPU is staying roughly the
same, and we're getting more and more capability by adding more, and more, and
more cores to a given CPU.

So on a given box there are many, many different things going on, and there's
communication amongst those little processors. In a data center, at the complete
opposite extreme, there are many, many computers operating in parallel, and there
are networks that are connecting those.

The programming tools, and the programming models that we as a company offer to
our developers, and we as an industry have to offer to developers, have to become
more parallel, both at the level of targeting one computer, or targeting multiple
computers in the data center. That's a lot of the innovation that we have to do in
our tools and languages group. Our operating systems are getting better, and better
in terms of the number of processors that they can schedule and manage workloads
on concurrently. Craig Mundie has a lot of work that he's been doing for a number of
years innovating in that realm.

Ultimately, some programmers won't necessarily program in a parallel way, but
they'll take advantage of services that use those multiple processors. They might
call a very, very simple library that will use the camera on a PC to recognize the
user, to understand a gesture that they might be using. We use the multiple
processors in recognition of handwriting, in recognition of voice, and so on. And that
doesn't require the programmer to get involved in those kinds of things.

CHARLES DI BONA: We're almost out of time, but maybe in the last couple of
minutes here, I think you've laid out a vision sort of where things are sort of headed
over the next five years, 5-10 years. Maybe you could sort of help from a financial
perspective, from a financial analyst perspective, what are sort of the things that we
should be looking for, sort of the tipping points that we're seeing progress, both for
Microsoft, but also for the market as a whole as it moves in this direction?

RAY OZZIE: I think it's -- again, I think the best way to look at it is really on an
audience by audience basis, from a market segment by market segment basis. On
the enterprise side, which is a big part of our business, I would not look in the next
few months, or even a year, in terms of specific material momentum, in terms of
people moving stuff into the cloud, but I would be very aware that this is a trend
that's actually happening, and we have strong backlog of customers who after we
announced our online initiative, are starting to work with us in a trial mode, to
experiment and to understand how they will manage these systems, how and at
what rate they will migrate users.

So I would just keep a close eye on how enterprise customers are, themselves,
balancing those uses of internal investment, investment in internal systems and
external. I would look at developer platforms. We've got a very significant
investment in developer models that let developers build things for multiple different
types of devices, and developer tools kind of give you an early indication of where
the solutions that will be built on those tools will be happening over the next few
years. So I think those are really the core DNA elements.

What really excites me about this whole thing is that the -- this constant change in
technology, in the nature of technology over time, really gives us, as an industry, an
opportunity for growth, and us as Microsoft an opportunity for growth. And services
in particular gives us an opportunity to take a number of the thing that we've done,
and to reach the next billion people, or the next 2 billion people that we might not
have been able to reach with the types of software that we've got. It gives us
opportunities to use the same assets we've got in different form factors, to take our
understanding of a value proposition, and to re-purpose it for people who have
slightly different needs, and different markets.

I think ultimately the opportunity for Microsoft is to create compelling, seamless
experiences that take the power of the Internet, and the magic of software, and to
mix them together, and to put them out there on a world of devices. And I would
just monitor those kinds of trends, in terms of devices, and backend infrastructure,
and how we're doing against those.

CHARLES DI BONA: Great. Thank you very much. We're right out of time.


To top