cirruslvr, <I think how the P4 does initially will depend on how the media (I don't mean Tom and Anand) spins it and how much systems cost relative to Athlon systems.>
Interesting that ZDNet is already quoting Mike Feibus of Mercury Research about the P4's low performance per MHz:
Pentium 4 is designed to be cranked up in terms of clock speed, which makes it difficult to compare directly to Pentium III or AMD's Athlon, Feibus warns. Benchmarks, which will shed more light on Pentium 4 performance, should be available at launch.
"We're not talking about performance, but clock rate. Clock rate is not a good measure of performance," he said.
zdnet.com
Spooner's article also has a timeline:
1.4,1.5 GHz - Nov. 20 release
o (moved up a week again?)
1.7 GHz or faster - "by the first quarter of 2001"
o I think this means by the end of the first quarter
approaching 2GHz - "by the second quarter"
o i.e., 2 GHz will occur in the 2nd half
This interpretation is consistent with:
The new 0.13 micron process will give Pentium 4 the boost it needs to move up past 2GHz clock speeds in the second half of 2001
and
Mid-year, Intel will begin the transition to a new manufacturing process, based on 0.13-micron interconnects. It's likely the first Celeron 0.13 micron chips will be mobile because the new process will allow Intel to increase clock speed, while reducing power consumption and costs.
o Notice that Intel is only beginning the transition to 0.13u output mid-year and the quote implies that the first 0.13u chips will be mobile chips, NOT P4 chips. This would be consistent with the way Intel rolled out 0.25u, but I don't remember if their first 0.18 chips were notebook chips or Coppermine.
It appears that getting to 2 GHz requires 0.13u and that 1.5 GHz will be the fastest speed grade until late March. It seems to me that Intel's only hope to overtake AMD is to reach 0.13u ahead of AMD by a significant amount.
Petz |
