Author: Bruce Moreland
Date: 01:20:12 08/11/99
Go up one level in this thread
On August 10, 1999 at 23:27:12, Pete Galati wrote: >On August 10, 1999 at 22:18:22, Tina Long wrote: > >>Hi guys, >>I remember Komputer Korner & others discussing this a year or so ago. >> >>I believe there was some concensus around the result approximating: >> >>"Once the hash tables are full, the time taken to reach a certain depth of >>search is 20% longer than if the hash table wasn't full." >> >>I'd appreciate hearing thoughts on this. >> >>Thanks >>Tina Long > >It's probably not the proper way to do it, but if computer starts making that >obnoxious chatter sound then I know I've got the hashtables set too high, so I >turn them down. For me it's that easy. I don't really care HOW much it's hurting >the performance, the point is that it IS hurting the performance, and besides I >don't want to listen to my harddrive. > >Pete She is talking about something else. What you are talking about is over-commiting memory, which is something you don't want to do in this case, it won't slow you down by 20%, it will slow you down so much you can't move the mouse around. What she is talking about is having the hash table set too small, so that what happens is that it fills up with entries, so new entries either have to overwrite old ones or not get written. I think this is implementation dependent. bruce
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.