In recent years, Sun Microsystems (SUNW) Chief Executive Scott McNealy was widely criticized for sticking too closely to Sun's longtime business model. Its model: Spend big on loads of in-house technologies, from chips to operating software, then try to sell products at a high enough price to cover the investment and still crank out a hefty profit.
He clung to the approach even as corporate buyers started snapping up cheaper machines based on Intel-style microprocessors that ran free open-source software, including the Linux operating system. Until 2003, McNealy kept on pushing only gold-plated systems featuring Sun's own processors, its Solaris operating software, and various other so-called middleware programs that are used to connect and deliver separate applications.
But even if investors still have reason to grouse about Sun's depressed stock price, they certainly can't say McNealy is stuck in the mud anymore. In February, Sun began offering free downloads of Solaris. In September, it announced innovative new servers built around Advanced Micro Devices' (AMD) Opteron chip.
RECAPTURING PAST GLORIES? The company claims the systems, designed by fellow Sun co-founder Andreas Bechtolsheim, offer far more performance while running much more coolly than rival machines (see BW Online, 11/14/05, "Sun: Back to High Performance").
And Sun has divulged a radical new chip design -- called Niagara -- that could provide similar benefits for a different class of server. Systems based on these chips, which should be announced by the end of the year, are designed to help big Internet companies such as Google (GOOG) or Yahoo! (YHOO) process millions of relatively simple transactions -- say, execute a search. That's in contrast to most processors, which were designed to handle hugely complex computing problems, such as forecasting the weather.
Maybe the most radical change of all came on Nov. 30, when Sun said it would give away almost all of its software products -- either individually or as part of a soup-to-nuts bundle called the Solaris Enterprise System. The goal is to rekindle the virtuous cycle that propelled Sun to greatness in the past.
RISKY MANEUVER. For starters, giving away the software could persuade more entrepreneurs and corporate programmers around the world to develop products to run on Solaris. Beefing up the developer base raises the odds that tomorrow's hot programs will run best on Solaris.
That could lift demand for Sun servers, as occurred in the late 1990s. Then, interest in Solaris and Sun's Java programming language lured buyers to its servers and gave Sun a chance to sell its other products as well.
And even if the software is free, big corporate buyers almost always buy the accompanying large service contracts to ensure they get proper customer support, such as bug fixes. This is how open-source software companies, such as Red Hat (RHAT), make money. The move puts Sun in sync with the way software is being developed, and with how customers want to pay for it.
Still, it's a huge risk. If Solaris doesn't gain market share or if Sun can't find enough ways to get customers to pay for services to cover the cost of software development, the move could backfire. Sun has yet to win back the allegiance of investors who exited its stock after the Net bubble burst. Despite all McNealy's model-busting announcements in the past few years, Sun shares still sit at just $3.88, down from $4.14 in early December, 2002.
McNealy is nonetheless hopeful. Famous for his outspoken opinions and humor-laced digs at competitors, he has adopted a lower profile since naming Jonathan Schwartz president of Sun in April, 2004. But there's plenty of fight, and fun, remaining in McNealy, as is clear from a Dec. 1 interview with BusinessWeek Computers editor Peter Burrows. Edited excerpts follow:
It has certainly been a busy news week for Sun, given the information that you plan to give away much of your software. Hey, we've had a busy year! You've got to give us credit for that. How many companies have been as interesting to watch as Sun over the past 18 months -- outside of Google, I suppose.
How can you make money on free software? Isn't 'We'll make it up on volume' one of the oldest formulas in the book for certain failure? Go ask {CEO] Eric Schmidt over at Google about that. Not too long ago, everyone was wondering, how can they do [all those searches] for free? But there's lots of ways to monetize [heavy usage of a technology]. And if we build a large Solaris community, who do you think is going to benefit the most? Look at Java.
Still, the market data indicate that Sun lost share in servers in the third quarter, while most of your rivals gained. Look at it this way. We've gone from No 99 -- nowhere -- to No. 4, or maybe even No. 3 in the last three years in sales of x86 (Intel-style chip) servers. And we're growing faster than IBM (IBM), Dell (DELL), and HP (HPQ) combined. No one disputes that.
And we've gone from 0 to 3.4 million downloads of Solaris since February. That is orders of magnitude greater volume than any previous generation of Solaris. That is huge.
Maybe so, but how can you be sure that those downloads are going to turn into real applications, ones that will drive volume for Sun gear? Programmers don't take the time to download a complete enterprise operating system unless they are going to do something with it. And when they do [create a program and want to run it on their corporate network], their boss is going to say: "Do you have a support contract?" And if it goes into production, the company is going to need some computers to run it on.
But can you make as much money per customer as you currently do? You mean, what's the razor and what is the razor blade? Well, the software is the razor. The razor blades are the servers, the storage, the memory, the service contracts, the archiving services, the tape cartridges, the integration, the consulting services. The whole deal.
Have you now made all the big changes required to put Sun back on the comeback trail so it can start gaining share and lifting its stock price? We've completely redone our product line and our strategy. What we need to do now is get the image of the company back to where it was. But we're making progress.
The noise around the company has changed. You haven't asked me: "Is Sun going to survive?? I haven't been asked that in 18 months. Now people ask: "Is Sun back?? Once people are comfortable with that, our sales cycles should get shorter. If we can show a little bit of sustained growth and profitability, we can turn this whole thing around.
So why do you think Sun is the only one of the big computer companies making such bold changes to its business strategy? We don't see IBM giving away its software like this. It happens to all of us. It happened to us at Sun. We were paralyzed by [the huge success during] the first phase of the Internet [in the 1990s], until someone came and hit us upside the head.
So then we made the move to Opteron, and to open source and to all the other things we've done. Well, Dell and IBM and some of the others are all thinking: 'Hey, we're on a roll.' But Sun is back. Those guys are in trouble, but people just haven't figured it out yet.
But again, it's so hard to know when that day is going to come. The company always has so many interesting technologies and bold initiatives, and you're always so optimistic. For example, you said earlier in this interview that "Sun is back.? Do you really think we're at that point in time? Have you said that before? Probably every day for the last five years [Laughing]. That's my job.
So far investors don't seem to be buying the story. Why not? We have to put up the numbers. But a lot of this is also related to what our competitors are saying about us. A lot of this is people kicking us while we're down, and I don't blame them. I'd do it, too. But we'll have our day in the sun.
READER COMMENTS
Burrows is BusinessWeek's Computers editor in the Silicon Valley bureau
Printer-Friendly Version