SemiconEng, yes, of course Intel is doing something about it. Given the war of the rumors over the past couple of years, I speculate that there are two opposing camps within Intel: x86-64 answer to Opteron vs. IPF answer to Opteron.
We now have some new information to digest - it is reported in the past few days that Intel will have the manufacturing economies of Itanium ready to replace Xeon around 2007. (Not just the manufacturing cost of Itanium, but the full cost of the motherboard with all the associated logic.) Now, three years out is hard to predict schedules, so I take this to mean that it will be about 5 years before Itanium can replace Xeon in the mainstream server market.
That should tilt the war in favor of the x86-64 advocates. Intel can not wait to the end of the decade to answer Opteron, and they know it.
I am finding the idea of a 2005 release of Intel x86-64 to be increasingly likely. I suspect it will be instruction set compatible with AMD64, but based on P4 technology. Probably won't sport an on-die memory controller, but will sport PCI-Express, a large L2 cache and iHT.
AMD has one year to establish themselves as the premier x86-64 provider. Paradoxically, the more obvious it becomes that Intel's version is coming, the more interest there will be in AMD's solution now.
Yesterday we watched a cat fight on the driveway. Lots of fur all over the place! Hard to figure out who won. Maybe an omen for 2005?
Market Data 
Markets 
AI
