Author: Robert Hyatt
Date: 10:36:26 10/24/03
Go up one level in this thread
On October 24, 2003 at 07:41:53, Gian-Carlo Pascutto wrote: >On October 24, 2003 at 01:16:21, Dann Corbit wrote: > >>On October 23, 2003 at 03:26:53, Gian-Carlo Pascutto wrote: >> >>>On October 21, 2003 at 22:33:55, Bob Durrett wrote: >>> >>>>On October 21, 2003 at 08:47:36, Bill Bahrenfuss wrote: >>>> >>>>>new SSDF List here http://w1.859.telia.com/~u85924109/ssdf/ >>>> >>>>This finally answers a question I asked months ago and got no answer, perhaps >>>>because no one knew. Junior 8 is better on a single processor machine than >>>>Deep Junior on the same single processor machine. >>>> >>>>Could it be that the Deep programs have overhead which reduces performance on a >>>>single processor computer? >>> >>>I can't speak for Deep Junior, but Deep Sjeng definetely runs at full strength >>>on a single processor machine too! >> >>A small difference in playing strength can be attributed to compiler flags for >>SMP verses single threading. The reentrant routines are sometimes slightly >>slower. > >Not for Deep Sjeng. > Your statement is wrong in context of his last comment. Thread-safe libraries _are_ slower. They _have_ to be. But not by a lot. And I would agree with you on the "not 10%" point. I have run SMP vs non-SMP crafty many times. It is about .1% difference, roughly. >-- >GCP
This page took 0.01 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.