To: pgerassi who wrote (140592 ) 7/31/2001 5:38:05 PM From: Tenchusatsu Read Replies (1) | Respond to of 186894 Pete, <"We didn't try real hard until after we acquired ...", isn't that saying we do not need 8 way until we could? Intel was implying that it was going higher than 4 way as a goal before that acquisition was made.> There's a difference between pioneers and copycats. Intel's foray into 8-way SMP was pioneering, not because they were the first, but because they brought standard high volume (SHV) economics to a space that was once dominated by expensive, proprietary solutions. Thanks to the pioneering efforts of Intel, there is a new market opened up to them. So yes, no one (or very few) wanted 8-way systems until Intel pioneered the way. This is a lot different than "No one needs it until AMD gets there," because the truth is that Intel is already in the areas that AMD wants to shoot for. That's a copycat, not a pioneer. To AMD's credit, they have the potential to pioneer HyperTransport into this market. And I'm not arguing against the benefits of HT. But I am arguing against your assertions that the 8-way market Intel currently dominates is no good until AMD drives HyperTransport into it. That's a load of FUD and bull. <Or how about claiming that a recompile shouldn't be done for Sledgehammer when it is exactly what you want later for P4 benchmarks> You are confused. I never said that recompiles shouldn't be done for x86-64. I only took issue with those who criticized Itanium and Pentium 4 for requiring recompiles but at the same time ignored the same requirement for Sledgehammer. Heck, if AMD can do it, they ought to get the software guys to optimize for Sledgehammer. My assertion is that they can't and they won't, that they'll continue to rely on their "Build it and they will come" strategy of (non-)marketing. Tenchusatsu