Author: Vincent Diepeveen
Date: 08:56:11 10/02/02
Go up one level in this thread
On October 02, 2002 at 05:55:39, Tom Kerrigan wrote: >On October 01, 2002 at 08:00:26, Vincent Diepeveen wrote: > >>Obviously the bitboard programs will profit. We can guess from alpha >>specs that it's about 33% profit. > >Really? How can we guess that? What specs? > >>A default compile from DIEP at an Itanium2/McKinley for example it is >>33% faster at it than it is at a K7 at the same clock. >>That's *very* good. Obviously it's a mixture of Level caches and especially >>doing bundles of 6 instructions a clock. > >What a load of BS. 33% faster than a 1GHz Athlon gives you 1.33GHz Athlon >performance. How is this "*very* good"? You can get a 1.33GHz Athlon on >Pricewatch for $52. > >-Tom Most supercomputer processors are stupid alpha's or stupid other cpu's which run a) at lower Mhz speeds b) give lower performance So i definitely find the McKinley a great cpu. It's obviously true that up to a processor or 8 you better can buy a dual k7 for $2000. Even if 8 processor mckinley is faster, you can let the dual k7 run a bit longer there. No problem... But now let's look it from supercomputing viewpoint. That McKinley CAN be put in most supercomputers. A k7 never will. So from supercomputing viewpoint the McKinley is great. 33% faster a Mhz. Alpha 21164 for example was a bad joke here. Despite some people saying it was ok for them, for DIEP it was horror. Sun didn't do too bad, but doesn't even get *close* to k7 speed for me. Also highest clocked Sun also is 1Ghz. So the McKinley is a big winner from all respects. Now let's discuss its price versus a SUN 1Ghz :) Best regards, Vincent
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.