Author: Mike Byrne
Date: 14:43:49 02/07/04
Go up one level in this thread
On February 07, 2004 at 17:33:37, Paul Doire wrote: >Hi , >This is with Mike's new compile. >542 Mhz, Win98SE >128MBRam, 16Hash >3,4,5 Tablebases >Ponder Off >Own Books >5 Min blitz games > >CraftySE(2) vs Ruffian 2 Own books Blitz 5' > > 123456789012345678901234567890 >1 Ruffian 2.0.0 01010½1½1001½101011001½0010011 15.0/30 225.00 >2 Crafty 19.10 SE 10101½0½0110½010100110½1101100 15.0/30 225.00 > >Crafty looks good here. Also Mike on my machine your newest compile >is faster than both Dann's and Peter's. It's hard to believe 10%-15% speed up makes a difference - but it seem to for Crafty against Ruffian. The latest PGO version is about 10-15% faster on all non P4 Intel processors. This includes P3s, AMDs and Celerons. For ALL P4s (so far) , the original version (dated 2/2/04) is faster. Ironically, the PGO version was compiled with Intel's Profile Guided Optimizations tool. this tool makes a faster compile for the non P4 (AMD, etc)) and a slower compile for P4s (Intel). I have a P4 (Intel). Based on the benchmarks received, Crafty is much faster on AMD on a per/GHz basis. In the old world (pre P4), that never used to happen. My next machine may be an AMD processor based.
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.