The Left and Right Hands of Sun

Published: September 6, 2007

by Timothy Prickett Morgan

Sun Microsystems hosted its annual financial analyst meeting in New York on Wednesday, and the two officemates who run the company--president and chief executive officer Jonathan Schwartz and chief financial officer Michael Lehman--suffered through BlackBerries interfering with their microphones for a few hours to calmly and rationally describe Sun's strategy for growing the company now that its finances have more or less been put into balance and it has a respectable product line again.

Getting Sun to grow again would be a whole lot easier if the whole Web 2.0 thing turned into a secondary dot-com boom, like we had in the 90s. But plenty of companies remember the IT hangover from five and six years ago, and none of them, with the possible exception of Google and a handful of companies that provision armies or refine oil, are going to be part of that kind of a bubble blowing operation again. So this time, in the Aughts, the economy is not roaring, business is tough but reasonable, and companies are trying to cope with a lot of change. It is no wonder, then, that Sun's top brass is excited to be on track for a 10 percent operating margin for fiscal 2009. A decade ago, Sun was bringing roughly that much dough, in terms of percent of revenue, to the bottom line. It was clearly much more fun to be running Sun a decade ago.

But, Schwartz, who has been at the helm for a year and a half, and Lehman, who was the bean counter back during Sun's boom years, too, are suited to the new challenging time that Sun faces. And in the more collaborative IT environment that Sun is trying to negotiate--in both senses of the word--Schwartz is arguably second to none in being open, and calmly extolling the virtues of openness. If anyone has been driving Sun's strategy of delivering free open source projects that lead to commercially supported products, it is Schwartz. Of course, this is a lot more boring than when Scott McNealy, Sun's founder, was calling the shots and often shooting his mouth off, but then again, the IT industry has consolidated and is in many ways a lot more boring itself even as IT has become ingrained into the very fabric of business.

"Our view is that, for the most part, we have been making really great progress," Schwartz said in his opening statements to Wall Street analysts. "I think you are going to have a really hard time finding a metric where we didn't make progress, with one exception. And this is the one exception that I and my team are making our focus, which is where do we find the growth?"

Sun's strategy is really quite simple, and Schwartz illustrated it in a set of foils that had technologies on one side and business opportunities on the other side--the left and right hands of Sun. And like Louis Gerstner did for IBM more than a decade ago with a certain amount of smarts (also a McKinsey & Company whiz kid, like Schwartz is), Sun's CEO is focusing research and development on Sun's core competency--operating system and system design--in such a way to directly lead to product and services. R&D is not an activity, but the foundation of specific product sets, which is how it should be anyway.

To that end, Sun is keen expanding its presence with the tens of millions of developers and administrators who have influence over IT spending, but who do not actually call the shots. Plenty of established technologies today--the Internet is a great example--were not productized until long after the nerds had been playing with them. But once the Internet became established, it basically wiped out every other public and private network. And that is why Sun is always talking about how many Solaris downloads it has (9.3 million as of last count), or how many Java downloads it has seen (over 65 billion so far). These are, in Sun's view, leading indicators for the products and services that will eventually lead to revenues and profits.

"Most vendors are not going to be paid for their software distribution," Schwartz explained. "In fact, we are going to see quite the opposite occur. Most vendors are going to be paying for distribution because they are worried that they are not going to get out into the marketplace." In that case, it is far better to give away a product and collect a service than it is to try to peddle a product for a licensing fee and then collect a 20 percent maintenance fee. Open source approaches to software delivery do away with the perpetual license and basically try to collect only the maintenance fee. To get the same amount of revenue from that business, you need to grow the installed base by a factor of five.

Sun is also counting on the transition from its legacy UltraSparc systems to new server, storage, and networking boxes based on Solaris and a mix of multicore Sparc and X64 technologies to turn around its systems business and get it growing again. Sun's "Galaxy" X64 and "Niagara" Sparc T1 businesses are now approaching an annual revenue run rate of $2 billion as of the final quarter of fiscal 2007, ended in June. Growth has slowed in recent quarters, and it is hard to say how much of the blame for that is due to Advanced Micro Devices' delays in getting its quad-core "Barcelona" chips to market or due to Sun not having long-since established a set of Galaxy products based on alternative X64 processors from Intel. You always have a second source for parts, as any IT supplier knows full well. Nonetheless, Sun has partnered with Intel and it is working to flesh out the Galaxy line with Intel boxes.

"As that growth begins to eclipse the legacy installed base that we have," Schwartz said, referring to products like Niagara and Galaxy in his charts, "we are fully expecting a return to more aggressive growth rates. While 6 percent growth is a good place to start, we would obviously like to see that number move up and to the right."

Schwartz also bragged a bit about hybrid products, like the "Thumper" storage server. The Galaxy X4500 is a two-socket Galaxy server with 24 SATA disk ports, which can deliver a gigabyte of capacity and high-bandwidth into the server for around $1.50 per gigabyte. Thumper has been for sale for two quarters, and is at a $100 million run rate, delivering 60 percent gross margins. "If this was an independent company, we would be looking at how do we take this business public," Schwartz said.

(That is what qualifies as a joke for Schwartz. I don't know what McNealy would have said in his place, but it would have made you actually laugh. Probably something like this: I got nothing against VMware, aside from the fact that Solaris containers are a better way to virtualize servers and a container doesn't cost three times that of the physical box it runs on, but give me a break, people. You could virtualize every stinking one of the 25 million servers on planet Earth with ESX Server, right now, and VMware would not be worth even half what it is worth on paper. Next time you guys on Wall Street are smoking something, do me a favor and share.)

Anyway, judging from the success of the public float of VMware, owned wholly by EMC and now with a market capitalization of $26 billion, maybe taking Thumper Storage Arrays Inc public is not a dumb idea, even if there is absolutely no rationality behind VMware's excessive valuation--excepting the flaws of human nature.

The point is, products like Thumper--based on Sun's iron, its Solaris Unix, and the related but not captive Zettabyte File System--herald the way that Sun is going to design products now. "Thumper's success is a derivative of ZFS, which has been advanced by developers who did not have any money to begin with. But if we did not amplify the developer community on the left hand," Schwartz said referring to one of his charts, "then the opportunity would have been considerably smaller on the right hand side," he said, referring to Thumpers feeds, speeds, revenue, and profit. "The more we proliferate technologies on the left, the more we amplify opportunities on the right."

Sun is, of course, putting Solaris at the heart of its systems, storage, and networking equipment, like it is doing with Thumper arrays and the giant "Magnum" InfiniBand switch that the company just launched. This is why I have argued that picking JAVA as its new stock symbol on the Nasdaq stock exchange was silly, and that Sun should have picked SLRS. And as Sun builds new server, storage, and networking technology, it is coming back to its root phrase, "the network is the computer."

"What we are really building are systems, where the traditional dividing lines between products go away," Schwartz said. "Our systems business is really the consolidation of these three very different businesses." The lines between competitors and partners are blurring, too, which is why Sun has partnered with Intel and IBM to get them to endorse Solaris on their respective iron. "The lines between who is a competitor and who is a partner are dissolving, and that is a deliberate part of the strategy."

In other matters, Sun also announced during the Wall Street meeting that it would be undergoing a four-to-one reverse stock split to get its share price above $20, which will not do anything to alter its market capitalization but which will, in certain cases, allow institutional investors to trade in the stock if they have lower limits on stock values they can trade in. (Apparently this is true.) The move undoes a bunch of stock splits Sun did in the 1990s as it was exploding. Sun also said that it would delay its reporting of its first fiscal quarter of 2008 financial results, ending in September, until November 5 as it takes some extra time to double-check the numbers in its new ERP system.

                     Post this story to del.icio.us
               Post this story to Digg
    Post this story to Slashdot