Author: Robert Hyatt
Date: 08:25:54 05/06/04
Go up one level in this thread
On May 06, 2004 at 02:50:19, Gian-Carlo Pascutto wrote: >On May 05, 2004 at 18:24:11, Robert Hyatt wrote: > >>I see 16-cpu speedups vary from 6.0 to 15.3. That is not "very small >>variability" if we are talking about the same thing. The numbers are all over >>the chart, just like current parallel numbers are with Crafty... 4-cpu tests >>were not nearly so variable, but that is the difference between DTS and the >>current less-effective algorithm... But even 4 cpu tests varied from 2.6 to >>4.0. Current Crafty would almost certainly produce 1-2 results > 4.0, why DTS >>did not is anybody's guess... > >Over half the of the 2 cpu speedup numbers are 2.0 exactly, yet none of them >is > 2. > >You truly are a man of amazing coincidences. > >-- >GCP Happens. I have hundreds of tests with no super-linear speedup from Crafty either. As for DTS, when you think about how I did the splitting, you can figure out why super-linear speedups were very unlikely...
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.