I'm not in the market for a EE or a Hammer. I'm not planning on more than a new videocard between now an whenever Tejas' get cheap.
If you had hopes AMD would catch Intel sleeping again like they did with the original Athlon, you were wrong. The EE only exists to protect the ultra-high end desktop market, and to strengthen Intel's place in low-end servers. I don't think anyone expects this chip to convert the AMDroids or to be a hit among the DIY enthusiasts. As Ed pointed out it will probably be a mediocre overclocker and it doesn't have a lot of bang for the buck. (At least at first-more later) But it will be enough to keep AMD from capturing the prestigious Falcon/Voodoo market.
These fanboy rantings are hilarious. Until a 64-bit OS and applications arrive, the FX's will be like driving a Ferrari Enzo in Manhattan. Microsoft says they'll have a 64-bit version of XP by December. Assuming they make a ship date for the first time in their 25 year history, then what? When are the 64-bit applications coming? Can anyone name ten 64-bit Windows apps that are scheduled to ship by next summer? 64-bit is going to make Mac look like a freaking Sam's Club of software. It is going to stay that way until Longhorn rolls out sometime in 2005 or 2006.
The EE will run on Canterwood and Sprindale mobos. This is going to give it a big edge in the retail channel over the FX. The 865 and 875 are stable chipsets with all the current generation features you could ask for. AMD is again putting its fate in VIA's hands. Even if the K8T800 ships without hardware conflicts and bug-loaded BIOS'es (about as likely as Microsoft shipping XP-64 on time) all the junk first-revision VIA chipsets that have shipped will give resellers all the reason they need to let someone else be the lab rat.
I don't see the EE cannibalizing IT sales from the Xeon. There have been lower-cost options to a Xeon for a while now. Yet the chip continues to sell well because it is part of a rock-solid platform. In a large enterprise even ten minutes of downtime can cost a lot more than the three thousand dollar difference between a Xeon and EE. Intel's dominance of this market won't change until someone can match the stability of the Xeon. (And it doesn't cost Intel anywhere near $3K to make a Xeon. That's what they charge because it is what their market is willing to pay. The profit margin on those chips is huge.)
But the EE will strengthen Intel's hand in the low-end server market. An EE can't compete against an Opteron running SMP or in 64 bit mode. But a small-office or mail server doesn't need SMP or 64-bit. In a single processor, 32-bit environment the EE will be a legitimate competitor to the Opteron, and this is where AMD planned to sell the majority of its CPU's. With clearly better performance AMD could charge a premium over a P4 server setup. But if the performance is comparable, Intel's reputation for greater reliability puts AMD back into a spot where it has to compete on price.
And that is the key point for the entire AMD FX/Intel EE fight. Intel doesn't have to beat the FX. All Intel has to do is offer comparable performance and that will force AMD to compete on price. That's exactly what they did. AMD had a chance to get off the mat with the FX and Intel drop kicked them in the face.
If that sounds harsh, it could get worse. Nothing would stop Intel from extending the EE range down the chart. If Prescott slips past December I'd be surprised if they don't launch a 2.8 and 2.4EE to jazz up holiday sales and raise its average unit price. A $700 3.2EE won't have much of an enthusiast market. But what about a $300 2.4EE, and how would AMD compete against that?
BHD