Author: Robert Hyatt
Date: 19:57:39 02/06/02
Go up one level in this thread
On February 06, 2002 at 16:07:36, athlon2000 wrote: >On February 06, 2002 at 15:57:11, Robert Hyatt wrote: > >>On February 06, 2002 at 13:47:30, Shep wrote: >> >>>On February 06, 2002 at 11:15:49, Robert Hyatt wrote: >>> >>>>On February 06, 2002 at 10:09:58, athlon2000 wrote: >>>> >>>>>Is shredders triple brain option any good to use against single chess program >>>>>users? >>>> >>>>I wouldn't think it is particularly good to use it _anywhere_. Would you >>>>prefer to have one engine thinking for X units of time, or three engines >>>>thinking for X/3 units of time and then choose what appears to be best? >>> >>>Just for the record, it's more like X/2 since the actual triple brain engine >>>does not seem to add much to the overall effort (more like some static >>>calculations, no search etc.), judging by the nps numbers. >> >>Does "triple brain" mean two programs + arbiter, or three programs? If the >>latter, it _must_ be 1/3 the speed since the programs are independent and >>will run on the same processor... > >professor hyatt it means two engines that run in parallel to each other while a >third engine observes the other two and decides which program has the better >move so in truth it would only be 1/2 strength so iff i where to use two strong >dual processors would a tripple brain be at full strength.. Or if you used a decent SMP program _it_ would also run twice as fast on that dual and the advantage you found goes away again and you are back to being 1/2 as fast... not a good deal.
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.