Author: Robert Hyatt
Date: 18:08:07 01/27/03
Go up one level in this thread
On January 27, 2003 at 05:57:12, enrico carrisco wrote: >On January 26, 2003 at 23:30:09, Robert Hyatt wrote: > >>On January 26, 2003 at 21:56:18, Omid David Tabibi wrote: >> >>>On January 26, 2003 at 21:44:20, Robert Hyatt wrote: >>> >>>>On January 26, 2003 at 20:02:05, Omid David Tabibi wrote: >>>> >>>>>On January 26, 2003 at 19:59:43, Peter McKenzie wrote: >>>>> >>>>>>On January 26, 2003 at 19:31:48, Omid David Tabibi wrote: >>>>>> >>>>>>>On January 26, 2003 at 19:28:39, Luis Smith wrote: >>>>>>> >>>>>>>>Full game >>>>>>>> >>>>>>>> 1.d4 d5 2.c4 c6 3.Nc3 Nf6 4.e3 e6 5.Nf3 Nbd7 6.Qc2 Bd6 7.g4 dxc4 8.Bxc4 b6 9.e4 >>>>>>>>e5 10.g5 Nh5 11.Be3 0-0 12.0-0-0 Qc7 13.d5 b5 14.dxc6 bxc4 15.Nb5 Qxc6 16.Nxd6 >>>>>>>>Bb7 17.Qc3 Rae8 18.Nxe8 Rxe8 19.Rhe1 Qb5 20.Nd2 Rc8 21.Kb1 Nf8 22.Ka1 Ng6 23.Rc1 >>>>>>>>Ba6 24.b3 cxb3 25.Qxb3 Ra8 26.Qxb5 Bxb5 27.Rc7 Line >>>>>>> >>>>>>>Yes, let me again repeat the repeated: "humans are still *far* stronger than >>>>>>>computers". >>>>>> >>>>>>You can say it as many times as you want but that doesn't make it true. >>>>> >>>>>Kasparov's performance does. >>>>> >>>>> >>>> >>>>The statement, on its face, is false. Had it said "Kasparov is still *far* >>>>stronger..." it might be true. but it said "humans", and as a collective >>>>class, that is simply false, and provably so... >>> >>>When we compare humans and computers, we are interested in the best from each >>>camp. Nobody is interested in the result of a match between Omid and Genesis for >>>example ;-) >>> >> >>No.. but the logical human "class" to consider is "grandmasters" and for >>that class, the computers don't have a long way to go to compete equally >>any longer... >> >>Deep Junior is running on purely ordinary hardware. My dual xeon is about as >>fast, based on NPS numbers. I can name a _bunch_ of GM players that would >>dispute the "humans are still far better" stuff. A few players might be able >>to "make it look easy". But the average run-of-the-mill GM will have to do a >>_lot_ of work to beat a computer, not that he can't do it... But the average >>GM is not enough better to not have to worry and sweat. > >NPS vs. different programs is no comparison (as you know...) Deep Junior's >1.6GHz x 8 is roughly 9.44 GHz after factoring in the SMP efficiency loss. This >is far greater than your 2.8 GHz x 2 box (including hyperthreading.) Maybe not. I have not run on a decent 8=way box yet. All suffer from severe memory bottlenecks. The quads are very good, but the 8-way boxes have all sucked badly when I have tested on them. IN fact, benchmarking suggests that an 8-way box is only 50% better than a 4-way box at the same clock speed. And when you factor in the SMP performance issues that a parallel search has (and nobody really knows how well DJ does at 4 or 8 processors that I have heard about) then the 8-way box begins to look worse. They are claiming 3M nps. The last time we played each other, our NPS rate was pretty equivalent. Doesn't mean it still is, but it is probable... > >The rest.... I agree with.. :) > >-elc.
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.