Did you go to serverworks website and actually see their products?
They make chipsets for Pentium III's and Xeons that are basically similar in design but somewhat beefier than the garden variety taiwanese chipsets that you find in a Dell desktop or a white box at CompUSA. The taiwanese variety don't have enough bandwidth for anything more than a desktop, so these guys make chipsets that have more bandwidth. They've already delivered "three generations" of products to the Intel world, presumably meaning that they started out with crummy old-style PCI and etc. and then oriented themselves toward the high end. Maybe they make good stuff--they have a long customer list.
Presumably, they are working on
a) Chipsets for the IA-64 series of boat anchors
b) Chipsets that will support the new I/O switch-based (rather than bus-based like PCI) architecture that everybody has known FOR YEARS would replace the current busses. Busses get saturated when processors get too fast and there are too many of them in a box. That's why everybody in the industry has been working on Infiniband FOR YEARS.
Where is the big deal in any of this whatsoever? What difference does it make if broadcom owns these guys or they're independent or whatever? They'll have competition, other people will make I/O chipsets for broadband and fiber and infiniband and other new I/O server things, on IA-64 and Sparc and Power and Alpha and whatever else. Sun will use somebody's chipset or make their own. Sun *already has* a much faster I/O system than the Intel guys. None of this is news. I really don't see how this is in the least significant. Maybe rudedog or some other disinterested third party tech heavy can comment.
twister, I haven't taken any steps back, I try to call 'em as I see 'em in both directions. You give Charles a hard time about viewing everything as bullish for SUNW, but you have never, ever viewed ANYTHING as bullish for SUNW even in the face of the company's huge growth and high-performing stock. This broadcom article is a case of making a mountain out of a molehill. That article says there are going to be better chipsets for IA servers. But there would have to be, or else there wouldn't BE any IA servers. Would there?
If you assert that there's something else I'm missing in that article, please explain it to me because I'd like to know. If your assertion is simply that when there's Intel-based competition for Sun then the competition automatically wins, that's already been proved false, and you're doing exactly what you accuse Charles of doing.
You should stop employing a double standard. If you want people to be humble, try setting an example.
--QS |
