Author: Joachim Rang
Date: 12:07:27 02/09/04
Go up one level in this thread
On February 09, 2004 at 14:59:42, Dann Corbit wrote: >On February 09, 2004 at 14:52:06, Joachim Rang wrote: > >>On February 09, 2004 at 14:47:59, Dann Corbit wrote: >> >>>On February 09, 2004 at 14:23:55, Aloisio Ponti Lopes wrote: >>> >>>>... by releasing their processors at the same speed in GHz ? >>> >>>Intel has the best fab in the world. How is AMD going to match that? The speed >>>is largely a function of the process and trace size. Intel's fascilities are >>>better. >>> >> >>that is only half of the truth (at best). AMD decided to concentrate on >>performance rather than on GHz. So AMD proc perform more instructions per cycle. >> I think it would be technically possible to release today a 5 GHz proc which >>performs only one instruction per cycle. The performance would be terrible, but >>one could claim it runs on 5 GHz. > >And you believe that a chip which performed 10 instructions per cycle would not >run at the same speed given the same trace size and process (e.g. copper >bonded)? > >I think if AMD had Intel's fascilities, they could produce Opterons at 3.4 GHz. > >I think many facets of AMD's design are better. But Intel's fascilities are >superior. No I think you are wrong. 10 instruction per cycle produce a lot more heat than 1 instruction per cycle. I think the design of the AMD procs dont allow >3 GHz yet (this will come of course with 90 and 65 nm). I don't think with the fascilities of Intel AMD would be able to release Opterons at 3.4 GHz but I may be wrong. Who is an expert in this field?
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.