Author: Vincent Diepeveen
Date: 11:24:57 09/03/02
Go up one level in this thread
On September 03, 2002 at 13:15:53, Matthew Hull wrote: I didn't look at his machine, i just looked to the search times Bob presented. Divided the 1 processor search time by the n processor data where n = 2,4,8,16 and then realize that the chance you get such great rounded off numbers is about 1/10 of a chance. just one number doesn't proof anything, but about 100+ numbers have a chance to be true: 1 / 10^100 = 0.000000000000000000000000000000000000000000000000001 where there are about 100 zero's behind the dot. That means in short it is a complete lie the search times. >On September 03, 2002 at 13:07:54, Vincent Diepeveen wrote: > >>On September 03, 2002 at 13:05:30, Matthew Hull wrote: >> >>It gives the reason why i took a closer statistical look >>at his data. In order to find what his speed penalties were >>for processors. > >I guess I don't understand how you know what the speed penalties are for a Cray. > Isn't that what the numbers are from, a Cray? > >>Then the fraud he committed was easy to find >>out. >> >>>On September 03, 2002 at 12:57:38, Gian-Carlo Pascutto wrote: >>> >>>>On September 03, 2002 at 12:54:05, Matthew Hull wrote: >>>> >>>>>Perhaps if you had a good understanding and experience of Cray architecture, >>>>>your statement would have more weight. But, the supercomputer you are using is >>>>>really very different from a Cray >>>> >>>>This has nothing whatsoever to do with whatever hardware was used. It >>>>has nothing to do with Vincent doing something different from Robert. >>> >>>Yes it does!!! look at the text from Vincent's post: >>> >>>"Nevertheless i worried about how to report about it. So i checked out the >>>article from Robert Hyatt again. Already in 1999 when i had implemented >>>a pc-DTS version i wondered why i never got near the speeds of bob >>>when i was not forward pruning other than nullmove." >>> >>>That's why Vincent is trying to discredit Bob's numbers, because Vincent can't >>>duplicate them. Looks obvious to me. >>> >>> >>>> >>>>The issue is the nature of the data that was presented in the DTS >>>>article by Robert. >>>> >>>>-- >>>>GCP
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.