Semi, Re: "I think though, that at 12 million in Q4, pretty much puts to rest that "dud" talk eh?"
Intel is great at making a screaming success out of a dismal failure. Who wasn't extremely disappointed when Netburst micro-architecture launched, and there were too many apps that couldn't show a noticeable performance improvement over a gigahertz coppermine? And not only that, but the processor was doomed to a failing memory standard, a dead-end socket, and costs that were much higher than their previous generation. Since then, however, clock frequency has improved 47%, and IPC has gained about 10%. Meanwhile, newer software is making much better use of the micro-architecture, costs have been significantly reduced, along with power requirements and die size, and memory support now includes all major technologies. Further, it is now on a stable socket that will last a long time into the future. All that sucked with the Pentium 4 has now been resolved. All Intel has to do now is evolve and proliferate the micro-architecture, and fill up the rest of the market segment holes.
Earlier last year, I was surprised that AMD didn't capitalize on the situation more. As Jim has been saying, they had the opportunity, but they executed like crap. This gave Intel the time they needed to fix their errors, and make their failures into successes. Now that they have a competitive product that stands up to anything AMD has to offer, there is no reason for AMD to keep the market share that they've gained. As Intel ramps up on fab capacity to meet new demand, AMD will have a harder and harder time clinging on to market share or revenue. Meanwhile, their only wild card, the K8, is still getting delay after delay, and no news of it has been broken since MPF months ago (when it had still not taped out). Who knows if AMD will ever get their act back together?
wbmw |
