Author: enrico carrisco
Date: 22:10:46 01/16/06
Go up one level in this thread
On January 16, 2006 at 18:44:52, enrico carrisco wrote: >On January 16, 2006 at 17:46:34, Sandro Necchi wrote: > >>I got the approval so I sent you what you asked. >> >>Check your e-mail box. >> >>Ciao >>Sandro > >Why would Shredder's speed on different platforms be top secret? > >Certainly this "within 10% of a Quad Opteron 2.4GHz on a Quad G5 2.5GHz Mac" is >not correct. > >I've already received the start position analysis from George Sobala who owns a >Quad G5 2.5GHz Mac and Deep Shredder 9.11 using 1 thread, 2 threads and 4 >threads. The SMP efficiency was excellent, but it was nothing to faint over and >certainly not more efficient than the Opteron's capabilities. With its >bidirectional hypertransport bus and super low latency it has incredible SMP >efficiency potential. What does the G5 have to match (or excel) this? > >I don't personally own 64-bit Shredder for the PC, but the native CB Deep >Shredder 9 on my machine (AMD Athlon X2 @ 2.85GHz per a core) is within 20% of >George's Quad Mac. Again, that is without 64-bit (which one would assume from >CEGT's tests and other sources that 64-bit would bridge the 20% speed gap -- or >come close...) > >So unless something "unexplainable" is happening from 2 threads to 4 threads on >the AMDs, a Quad Opteron 2.4GHz using Deep Shredder 9.12 should be 27-31% faster >than the Quad G5 2.5GHz Mac. > >I will have _EXACT_ numbers as soon as an associate of mine returns the >eval/analysis I asked for from his Quad Opteron system. > >Regards, > >-elc. I guess I was too conservative with my initial estimates... It's actually 35.8% faster. :) Full information here: http://www.talkchess.com/forums/1/message.html?480206 -elc.
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.