Online Extra: Sun's New Software Horizon
When Sun Microsystems (SUNW ) unveiled Java in 1995, programming languages were considered a backwater of the software industry. All the excitement surrounded the newfangled Web browsers hitting the scene. It turns out, though, that Java has been every bit as important. It freed software developers to write Web applications that run on any computer.
This sparked an immense wave of innovation that continues to this day. For all its influence, Java never made much money for Sun. But it helped position it as an Internet pioneer.
Sun CEO Scott McNealy is betting that the latest item from his software skunkworks, dubbed N1 (for Network One), will make an even bigger splash. The software is his version of high-tech utopia: Hundreds of computers working together as one big, easy-to-use machine.
TIPS FROM MA BELL.
McNealy wants to do away with the maddening complexity of a computer network, which requires legions of professionals to install software and corral idle horsepower. McNealy says with N1, the computers themselves will handle much of that work. How? They'll be programmed not only to know what's wrong -- whether it's a shortage of computing power or a software glitch -- but also how to search the network for a solution.
McNealy's model for simplicity is a phone network. "Remember when the telephone industry had switchboards and operators?" he asks. If the computer industry doesn't come up with good automatic systems, "we'll be forced to train everyone under the age of 30" to run networks.
This vision is hardly a panacea for Sun's ills. For starters, the first pieces of N1 will hit the market by January, but the technology that pulls it all together is still two or three years away. What's more, N1 could actually weaken hardware sales, which account for 63% of Sun's revenues and an estimated 65% of its gross profits. That's because N1 promises to make networks superefficient by optimizing the use of gear that's already installed -- a recipe for buying fewer computers, not more.
BEHIND THE CURVE?
En route to N1, Sun will face bigger and richer rivals. IBM (IBM ) is putting the muscle of its $5 billion research and development budget behind similar technology and has made it top priority. "This is probably one of the biggest computer-science problems you could find," says IBM CEO Samuel Palmisano.
And down the road from Sun, Hewlett-Packard (HPQ ) already has a similar product and 10 customers in pilot projects. "Sun is way behind at this point," says Ahmar Addas, managing director at Grid Technology Partners, a high-tech consulting firm.
Clearly, Sun isn't racing ahead of its rivals as it was with Java. But that's not stopping McNealy. He has spent at least $200 million on two acquisitions, picking up one technology that searches the Net for more computer horsepower and another that manages data storage over the Web.
Analysts believe McNealy will dip into his $5.2 billion in the bank to buy more. One gap: software that allows a network manager to see what each machine is doing -- a feature HP and IBM already offer.
If McNealy's gambit pays off, Sun could emerge at the forefront of new technology that revolutionizes computer networks. This time, however, instead of Sun leading the pack, it's part of the field.
By Jim Kerstetter in Sunnyvale, Calif.