Author: Robert Hyatt
Date: 19:13:00 12/11/02
Go up one level in this thread
On December 11, 2002 at 21:14:35, Aaron Gordon wrote: >On December 11, 2002 at 20:58:09, Robert Hyatt wrote: > >>On December 10, 2002 at 17:21:37, Matt Taylor wrote: >> >>>I would appreciate other people posting results as well for comparison. It would >>>be nice to have results localized in a single thread without lots of extra >>>discussion. >>> >>>-- Dual-AthlonMP 1600 (133 MHz FSB) -- >>>Total nodes: 121662897 >>>Raw nodes per second: 1382532 >>>Total elapsed time: 88 >>>SMP time-to-ply measurement: 7.272727 >>> >>>-- Dual-AthlonMP 2000 (133 MHz FSB) -- >>>Total nodes: 118211187 >>>Raw nodes per second: 1597448 >>>Total elapsed time: 74 >>>SMP time-to-ply measurement: 8.648649 >>> >>>-Matt >> >>Total nodes: 63863813 >>Raw nodes per second: 1030061 >>Total elapsed time: 62 >>SMP time-to-ply measurement: 10.322581 >> >> >>Total nodes: 67107793 >>Raw nodes per second: 2097118 >>Total elapsed time: 32 >>SMP time-to-ply measurement: 20.0000 >> >>The first number is the current version using one cpu only. >> >>The second number is using two 2.8ghz xeons. Run four times and >>averaged. > >Hyatt, are you also using 18.15? I can understand a small node variation but >~120m nodes vs ~65m nodes is pretty massive. The single CPU result seems on-par >with an AthlonXP 1900+ (1.6GHz). This was with Crafty v18.11 though.. If you can >please get some Crafty 18.11 results. No hash, just load up 18.11 w/o a >crafty.rc file and type "bench". All results below are single CPU results except >for the dual celeron mentioned at the bottom. No... it was not 18.15. I had written "this is the latest version" but I somehow deleted that when cutting/pasting the actual output. I'll try to compile 18.15 and test it... Here is the same test for 18.15. Machine = 2.8ghz dual xeon, 1.0gb DDR RAM, one processor without SMT and two processors with SMT enabled. Total nodes: 97487547 Raw nodes per second: 1107813 Total elapsed time: 88 SMP time-to-ply measurement: 7.272727 Total nodes: 118365893 Raw nodes per second: 2156433 Total elapsed time: 59 SMP time-to-ply measurement: 10.666667 Again the last run was averaged over four separate runs of the bench command, for those that like to forget that I do this most every time. > >XP 1.6GHz (1900+) >Total nodes: 104161030 >Raw nodes per second: 1001548 >Total elapsed time: 104 >SMP time-to-ply measurement: 6.153846 > >XP 1900+ @ 1.86GHz(~2200+) >Total nodes: 104161030 >Raw nodes per second: 1197253 >Total elapsed time: 87 >SMP time-to-ply measurement: 7.356322 > >XP 2600+ @ 2.52Ghz >Total nodes: 104161030 >Raw nodes per second: 1578197 >Total elapsed time: 66 >SMP time-to-ply measurement: 9.696970 > >and the dual Celeron 400 @ 550 comes in around 550kn/s.. poor lil' thing :)
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.