When Sun Microsystems Inc. (SUNW ) Chairman and CEO Scott G. McNealy took the stage at a shindig in San Jose's Tech Museum of Innovation on Nov. 15, he unveiled the most important product to come out of his company in nearly a decade. Called Solaris 10, it's the latest version of the server maker's flagship operating system and, by most accounts, an impressive piece of technology. Built over four years at a cost of $500 million, Solaris boosts Sun's technological lead on Linux by increasing the efficiency of the server computers it runs on and making them as heavy-duty as an old mainframe.
But will the new Solaris be enough to light a fire under the long-suffering Santa Clara (Calif.) company? In recent years, Sun has lost customers to low-cost rivals that piece together servers with chips made by Intel Corp. that run on the Linux open-source operating system. That's why cutting-edge software alone won't get the job done. It will take more effective cost-trimming and an improvement in Sun's oft-criticized sales force to convince investors Sun is back on track. "This is going to be a slow process that probably has a lot of bumps along the way," says Laura Conigliaro, a Goldman, Sachs & Co. (GS ) analyst.
McNealy's rebound strategy reverses Sun's go-it-alone model -- build and sell everything from the computer chip to the operating system -- that has cost the company dearly. Solaris 10 will work with not only Sun servers but also rival boxes and chips. That could hurt Sun server sales, but McNealy has no choice if he wants to keep buyers who like the cheaper, more flexible technology of open-source rivals.
In an effort to convince the marketplace that Sun has changed its approach, McNealy is determined to enforce a cultural shift among his sales force, which had long pushed customers to buy both Sun software and servers. Now, if a customer buys Solaris but runs it on a rival server, the salesperson will be compensated as though she had sold the box, too.
What's more, with Solaris 10, Sun has all but adopted the pricing model used by top Linux distributor Red Hat Inc. (RHAT ) When Solaris ships at the end of January, it will be free for use on lower-end servers.
So how will Sun make money? Like Red Hat, it aims to sell monthly subscription services for things like bug patches and support. For a typical server, that will cost customers about $1,400 a year. With traditional software sales, customers pay for everything up front. By contrast, Sun will likely see sales on a monthly basis. That should smooth out earnings.
The pricing model is an effort to stop customer defections to Linux. As Sun President Jonathan I. Schwartz acknowledges, the bulk of the $2 billion annual market for servers that run on Linux, up from near zero five years ago, came at Sun's expense. Sun's share of the $45 billion server market has dropped 5.5 percentage points since 2000, to 12.5%, according to researcher IDC. IBM, whose servers run its own OS and Linux, leads the market with a 32% share.
It won't be easy to get customers back from Linux' clutches. Switching operating systems is notoriously difficult. Zachary A. Nelson, CEO of Web-services company NetSuite Inc., won't buy Solaris. "It's difficult to go back in there and monkey with the operating system," says Nelson. He's sticking with Linux. Yet Sun's newfound flexibility has an upside: Nelson plans to replace his 300 Hewlett-Packard servers with Sun boxes.
Given Sun's low-price approach to the market and the continued threat from low-cost rivals, keeping expenses down is crucial. McNealy figures he can reduce Sun's costs by $500 million in the fiscal year ending in June, 2005, via layoffs already under way and cheaper computer-component prices.
Will it all be enough to turn Sun around? After three years of decline, Sun has had two quarters of single-digit sales growth. And in the quarter ending Dec. 31 the company is expected to generate $87 million in net income, although much of that will come from cost-cutting. It could be a year or more before investors know whether Sun is in the middle of a hard-fought return -- or a tech giant that couldn't change with the times.
By Jim Kerstetter in San Mateo, Calif.