Author: Robert Hyatt
Date: 10:47:59 05/06/04
Go up one level in this thread
Go here: ftp.cis.uab.edu/pub/hyatt/smpdata There you will find 3 logs from the 24 CB positions in the DTS article. one for mt=0, mt=2 and mt=4. mt=2 produced a 1.89 speedup. mt=4 produced 2.97. I believe my rough approximation speedup = 1 + (NCPUS - 1) * .7 is _very_ close to fitting a straight line through those two points. underestimates mt=2 by .2, overestimates mt=4 by .1 That is _really_ worth arguing about.
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.