All Intel investors
For all but the most finicky of technology buyers, the difference between SPARC and the Intel architecture (IA) is now, more than ever, about price. It's a war Sun cannot win.
The most obvious move for Sun would have been to make functional equivalents to Solaris for SPARC available for IA, on the same timetable that it makes them available for SPARC.
Even so, such a move (which my sources at Sun tell me is nearing reality) would hardly be enough to give Sun the rebirth that I believe it desperately needs.
In fact, the latest idea is to break with the Sun tradition that has long exalted Solaris/SPARC as the ruling technology and replace it with the heir to that throne: Java.
AMD is giving it a go with Opteron, trying to lure developers away from the IA, a move that I think will ultimately fail.
techupdate.zdnet.com
REALITY CHECK
Sun bets its future on Java
By David Berlind
August 26, 2002
Sun's recent moves towards supporting Linux might feel like a warm embrace for the open source movement. But it has much more to do with supporting anything that will enhance the Java ecosystem.
Sun CEO Scott McNealy has a cadre of advisors that's akin to Bush's cabinet. If there were such a job as Secretary of Technology, Sun Chief Engineer Rob Gingell would probably have that title. Within McNealy's cabinet, the 17-year Sun veteran has been a catalyst for change, and probably not a very popular one.
Four years ago, Gingell started to advocate the idea of open sourcing Solaris, Sun's version of the Unix operating system. The very idea must have been considered heresy. For Sun's entire history, Solaris and the hardware it's married to--the SPARC microprocessor--have been the crown jewels of the company. You can imagine how the mere suggestion of open sourcing one half of the franchise might have been looked upon as ludicrous.
In my interview with Gingell, he said, "Back when I first talked about giving away Solaris, it was about four years ago during one of the periodic 'IBM talking to Sun about using Solaris' conversations. And everyone here was like, 'GASP! IBM will see everything--including the secret sauce!'"
But now it looks like Gingell is about to get his way. Contrary to popular belief, Sun's recently announced forays into desktop and server-based Linux are not all about belatedly jumping on the same bandwagon as most of its competition (including IBM, HP, Dell, and Intel). Instead, the move is all about placing the biggest bet in Sun's history.
I've repeatedly maintained in this column that processing power--a.k.a. MIPS--is a commodity. For all but the most finicky of technology buyers, the difference between SPARC and the Intel architecture (IA) is now, more than ever, about price. It's a war Sun cannot win.
Sun clearly needed to swallow its pride and dive into the Intel market with both hands and feet (not just with token participation via storage appliances). The most obvious move for Sun would have been to make functional equivalents to Solaris for SPARC available for IA, on the same timetable that it makes them available for SPARC.
Even so, such a move (which my sources at Sun tell me is nearing reality) would hardly be enough to give Sun the rebirth that I believe it desperately needs. Gingell, and now much of Sun, apparently agree with me. In fact, the latest idea is to break with the Sun tradition that has long exalted Solaris/SPARC as the ruling technology and replace it with the heir to that throne: Java.
To accomplish that transition, Sun is doing more than acknowledging that MIPS are a commodity. According to Gingell, even the operating system is a commodity. From his point of view, neither SPARC nor Solaris are differentiators. Solaris doesn't have a lot of secret sauce, just good code, Gingell said. Instead, Gingell wants to see the parts of Solaris that can't be found in Linux open-sourced and then merged with Linux. He calls the result "Linux by Solaris."
Doing so would pave the way for Sun to follow its competitors and change the primary ecosystem around which it bases its business.
For companies like Sun, ecosystems are three-legged stools. One leg is developers, the second is applications, and the third is volume. Remove any one of the legs, and the stool falls over. For a long time, the Solaris/SPARC stool stood on its own three legs, thriving on a vibrant developer community, support for nearly every Unix application, and an enviable volume of server sales. In acknowledging that processors and operating systems are commodities, Sun--or at least Gingell--is conceding that the Solaris/SPARC duo has plateaued, and that it's time to move on to the Java ecosystem.
Some of Sun's competitors have already bitten the ecosystem-transition bullet. Intel started down this path in the early 90s when it began collaborating with HP on the EPIC foundation that underlies the IA-64 processor family. Microsoft is now in the process of erecting a whole new ecosystem around .Net. AMD is giving it a go with Opteron, trying to lure developers away from the IA, a move that I think will ultimately fail.
The Java ecosystem may be Sun's best and most natural bet. According to Gingell, Java may already have ten times more developers than Solaris. Despite Microsoft's recent decision to stop bundling the virtual machine with Windows, Java has a volume rivaled only by Windows. And the Java applications base is rising rapidly.
But it's also the stool's weakest leg. For Java to succeed Solaris/SPARC in the Sun kingdom, Sun had to concede that Solaris and SPARC are commodities. That would clear the way for Intel and Linux support, both of which Gingell also sees as commodities, and both of which are already a part of the Java ecosystem. In fact, given Java's independence from the operating system and hardware layers, Intel's and Linux's volume make them more important drivers of the Java ecosystem than either Solaris or SPARC will ever be.
The only question that remains is, how will Sun profit as a company? Sun doesn't have a lock on all things Java. Plenty of other companies--including IBM, Oracle, and BEA--lead Sun in areas like application servers.
It's also widely believed that the royalties Sun collects from Java and the fees it collects from certifying third-party software for Java compatibility are not enough to sustain the company's growth.
The solution, according to Gingell, is simple. "Despite popular belief, our business has always been about solving customer problems by building solutions on top of fairly standard things and just doing a better job of it than our competitors. SPARC is an IEEE standard. Unix isn't substantially different from one version to the next. We just did it better and faster than anybody else. When it comes to Java, there's no reason we can't do the same."
He may have a point. But Sun hasn't always had to compete with a 900 lb. gorilla named IBM on its back, a nemesis that will do just about anything to squash Sun and wrestle away control of Java . Winning this battle--or even coming in second (which Gingell says is O.K.)--will be significantly more challenging than winning the Unix wars. For customers, it should mean that the cost of commodity hardware and OSes will decline and that reliance on Java will accelerate. |
