Q&A: Sun's CEO points to revenue gains in servers and storage as proof of the strength of the OEM's product road maps.
Sun Microsystems is feeling something that it hasnt felt in several years: momentum. In the second quarter of 2006, the Santa Clara, Calif., company saw server revenue gains
of 15.5 percentthe only top-tier OEM to see revenue growthand moved past Dell to take the No. 3 spot among global systems vendors, according to market researcher IDC, in Framingham, Mass. Sun also is making strides in the storage space, recently rolling out its "Thumper" server/storage hybrid.
Internally, the company continues to undergo radical change. Sun announced in May a plan to slash its work force by 13 percentabout 5,000 jobs
and founder and longtime CEO Scott McNealy handed over the reins to Jonathan Schwartz.
On Sept. 13, Schwartz and other Sun executives will be in New York talking to Wall Street analysts about the companys future plans and giving them the message that, after several disappointing years, its back on track.
Schwartz spoke with eWEEK Senior Editor Jeffrey Burt about the changes at Sun, its products and its future.
Can you give us your thoughts on your first 100 days or so in charge of Sun?
The first thing that we did was I and the new team pulled together a really exhaustive analysis of all of our [R&D] activities, our market allocationsso, how we put our sales force and marketing dollars to workand then just the overall cost structure of the company, so what we could put together was our best view on how we achieve some measurable and discreet targets, which we then turned around in the first 45 days and articulated to the street.
What we said was, we want to grow the company, we want to grow the profitability of the company and we want to drive to four points of operating marginbasically four points of profitin the fourth quarter of this fiscal year.
Longer term, we actually said we think the market is heating up and coming our way. To that end, we articulated a longer-term objective of hitting 10 percentage points of operating profit and, again, trying to put some meat on the bones. Back when we made the restructuring announcement, I think it was interesting to people, but really became kind of declarative to them when the IDC numbers came out and showed just demonstrable, irrefutable share gains in a [server] market in which our competitors were really suffering.
That was just amplified by some discreet share gains on the storage side, and again, I dont think most of the analyst community really knows how to count some of the emerging trends yet, but weve seen demonstrable and irrefutable evidence that Solaris as well is growing, and again, for us, thats the flywheel in our business and thats whats creating all these new and emerging opportunities, both in the development community as well as in the IT and deployment community.
Sun brings JRuby in-house. Click here to read more.
So we made some pretty sweeping changes inside of Sun, driving new alignments, driving real systems innovation. Systems for us is the intersection of computers and storage devices and software platforms and service platforms.
As we drive towards that evolution in our business, we really see the market turning towards those who offer whole systems and not just a cheap box or a piece of free software or a storage disk. They really want the problem solved and we think we are someone uniquely qualified to go solve the problem. And again, now weve got some data behind that. Its not just our articulating a captivating vision of the future.
In making the changes over the past three months, had you anticipated being able to leapfrog over Dell for the No. 3 spot so quickly?
If you go back and read our press releases, what we have been saying is the adoption of Solaris is the leading indicator of our server business. By all means, we knew a year ago we were gaining steam
on IBM, [Hewlett-Packard] and Dell, but no one believed us when we said that the adoption of Solaris was the perfect leading indicator.
When Q4 turned around, we were probably surprised at the magnitude of the distance we put between ourselves and the competition, but again, we knew that Solaris was giving us access to customers that either wanted to move away from their proprietary Unix architectures, or wanted to move away from vendors who could only come in and say, hey, Ive got a really cheap price for you.
Again, we continued to hear of that momentum in the marketplace, and not just only from our server customers, but also from our storage customers.
To be honest, I was a little surprised at the strength of adoption of the Niagara platform.
For the first time, weve exceeded $100 million in one quarter. I was asked by a Wall Street analyst
who said, That was only one deal, wasnt it? That was Google buying $100 million of Niagara. And I said, No, those were trial units. And to understand the magnitude of the momentum we have in the marketplace, no ones going in with large-scale deployments of Niagara. So with $100 million in business, we were just planting the seeds in the field.
The other interesting thing is, whats going to happen when those seeds begin to sprout?
Was that $100 million primarily selling into your installed base?
Sixty percent of our try-and-buy units were going to new customers, so the majority were new customers of those trial units. I was with one customer yesterday who said he just replaced five IBM xSeries boxes with the Niagara platform. Were seeing vendors moving away from box-only solutions, whether theyre x86 or as well as SPARC, and
weve also seen a very aggressive adoption among the PA-RISC community now that HP, for all intents and purposes, has abandoned them.
When you look at the SPARC platform for Sun, how do you envision it evolving, especially in light of your other projects, such as Opteron and Niagara? For example, you recently took the "Serrano" [low-end UltraSPARC IIIi+] chip off the board, deciding instead to focus your energies on Niagara. What does this say about your plans for the evolution of the traditional SPARC line?
The reason we moved away from the chips that werent doing well is because we basically saw such a massive uptake in Niagara. Our view was, Why distract ourselves with stuff that isnt going to be revolutionary? To understand the magnitude of growth, we went from zero to $110 million in six months. So I am thrilled with the idea that we are going to move away from just incremental innovation and move toward revolutionary innovation when you see that kind of growth across the board.
On the high end, I think were definitely seeing growth. And again, HP shrank because they end-of-lifed [PA-RISC]. IBM shrank with [Power5] and P5+, and we gained share not only with the low end with Niagara, but also in the high end with UltraSPARC IV+.
We are ganging up our resources inside Sun on the winners. We are actively disinvesting in those areas that we dont think are going to yield the kind of growth that we want to see. Single-digit growth is nice; double-digit growth is a lot better. And weve got a lot of it.
What other areas are you looking to disinvest from?
Id rather focus on where were investing. We are increasing our investment in the Solaris platform and in our engagement with the developer. We are increasing our investment in the open-source and open community engagements, and thats on hardware as well as software. Again, Niagara is available through open source now. If you go to opensource.net.
Theres no other non-x86 chip out there that has the breadth of Linux support that SPARC does now.
We are actively disinvesting in proprietary technology, and in closed-source software development. [We are] actively looking to disinvest in areas that dont, for example, leverage Solaris in their core. There were a lot of systems inside of Sun that were running an alternative operating system or were not running our server infrastructure.
To me, another good example of that is were going to disinvest in building our own proprietary storage. Were going to continue to rely on Hitachi and, potentially, EMC. We just announced a new product called Thumper. Were going to be pouring a huge amount of resources and effort into driving Thumper across the world, because to me, the future is about NAS [network-attached storage] and tape. Its the bookends of the market that are really going to show a lot of the value. SANs [storage area networks] play a very important role in the marketplace, but the real growth youre going to see is going to be on the bookends.
Why are we doing that? Because it allows us to really leverage being a systems company.
We can build almost anything out of [Solaris running on a Sun Fire system] from blades to rack-mounts to workgroup to SMP [symmetric multiprocessing] systems to horizontally scaled systems, storage and servers. Plus its all about systems integration, where you bring all that to bear.
As we go forward, you should expect to see much, much more coordinated innovation. You will see the evolution of security technologies not only in Solaris, but in the Niagara chip and in our StorageTek tape offerings. They will coordinate. Why? Because I dont know about you, but Ive got customers who care about secure software, about secure servers and secure storage,
but what they really care about is security. So we want to be able to present to them an umbrella platform for innovation and not just point products.
What Im seeing in the market is that the growth margin in point products is just plummeting. Whether its free software or a one-way x86 server, nobodys going to make money on the components. The only companies that are going to make money are those that lash it all together in a solution that the customer finds appealing.
Possible partnerships for Sun.