Author: Chessfun
Date: 22:10:00 04/28/00
Go up one level in this thread
On April 29, 2000 at 00:49:56, Robert Hyatt wrote: >On April 28, 2000 at 09:41:29, Chessfun wrote: > >>On April 28, 2000 at 09:21:59, Enrique Irazoqui wrote: >> >>>On April 28, 2000 at 08:59:27, Chessfun wrote: >>> >>>> >>>>All games on one Cel 433. >>>>Ponder=on. >>>>Tablebases used at 25 and longer. >>>>Nunn 1 positions. >>>>No opening books are loaded. >>>> >>>> 1 min game Fritz 6a 10.0 - 10.0 Crafty 17-10 >>>> 2 min game Fritz 6a 12.0 - 8.0 Crafty 17-10 >>>> 3 min game Fritz 6a 16.5 - 3.5 Crafty 17-10 >>>> 5 min game Fritz 6a 16.0 - 4.0 Crafty 17-10 >>>>10 min game Fritz 6a - Crafty 17-10 running >>>>25 min game Fritz 6a 7.0 - 6.0 Crafty 17-10 Current score >>>>60 min game Fritz 6a - Crafty 17-10 Unplayed >>>>Tourney times Fritz 6a - Crafty 17-10 Unplayed >>>> >>>>These results seem to be all over. >>> >>>Since you are using one machine and ponder on, each engine gets 50% of CPU time >>>and therefore you should compare these results with the ponder off results at >>>half the time. 2 minutes game ponder on compared to 1 minute game ponder off, >>>etc. >>> >>>There is still the question of the CPU time being allocated properly. I wonder >>>if we are proving anything like this. >> >> >>I think ponder=on on one cpu just don't work. >>IMHO that is :-) >> >>Thanks. > >I think it works just as well as ponder=off. I don't know of any O/S, >from windows 98 to NT to any flavor of unix, that won't give almost exactly >50% cpu utilization. > >IMHO it is better than ponder=off as that is an unnatural mode for a program >to run in. two machines would be even better, but one works fine just so >you don't start turning this off, or that off, etc. It's a long story but I have two Celeron 433's each with 128 mb ram, one is running win 98 the other win 95. In the ponder=on that I ran I used the win 95 machine for all games excluding 25/0 which ran on the win 98 machine. But now it seems like autoplayer will work but has some bumps, two stopages so far for no apparent reason, both at the end of a game. I think the cause is simply loading the starting positions which it seems to do......master moving then slave loading the position. But it is easy enough to continue from the point left off by simply making new database of the missing games. I expect this will take longer than the one machine games as simply loading it to play is a little consuming and the load time and lag between games is much greater than on one machine. Thanks. > > >> >> >>> >>>Enrique >>> >>>>As anyone using F6 knows with ponder=on you get a question, >>>>Permanent Brain on single processor?. >>>> >>>>Thanks
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.