Author: Tom Kerrigan
Date: 20:02:08 06/29/00
Go up one level in this thread
On June 29, 2000 at 21:42:59, Albert Silver wrote: >On June 29, 2000 at 13:25:19, Tom Kerrigan wrote: > >>On June 29, 2000 at 03:38:38, Albert Silver wrote: >> >>>On June 29, 2000 at 02:21:10, Gregor Overney wrote: >>> >>>>>Not even close yet. That hardware would be approximately 1% of the power of >>>>>the DB hardware. And that is being _very_ generous... >>>> >>>>1% is a pretty good estimate for a four processor machine using four P5-4 >>>>running at 1.5 GHz using a four channel RDRAM bus that delivers 3.2 GB of data. >>>>Estimate 500 kNodes per CPU times 4 = 2M Nodes = 1% of DB's avarage performance. >>> >>>If one CPU achieves 500k nodes, I doubt very much that 4 CPUs will achieve 2M >>>nodes, unless 100% efficiency has been achieved. Crafty is apparently the most >>>efficient at this level though only Bob would be able to say how well it should >>>do. >> >>No, believe it or not, Bob Hyatt is not the only competent chess programmer in >>the world. > >:-))) Clearly you are trying to trick me Oh Evil One! But I am not fooled by >your daring attempts to lure me into your dastardly plots! > >Seriously though, all I said was that as far as I knew ("apparently") Crafty was >the most efficient multi-processor chess program among the micros. This wasn't >based on any worshipping of the 'Great One', but on what I had been led to >understand from my reading here. If there is a microcomputer chess program that >makes better use of multi-processor systems, please just say which. AFAIK, Crafty uses a fairly common/simple MP algorithm. It's possible that other micro programs are not using a better algorithm, but I doubt they're doing any worse. >>In any case, who says that DB was searching at 100% "efficiency"? > >Actually, I wasn't talking about DB's efficiency at all, but merely the >well-known figure of 200M NPS. Greg had said that a four processor machine would >reach 1% of DB average performance, and said this by merely multiplying the NPS >achieved by each processor. I questioned this conclusion and waited to be >corrected, that's all. If you're just talking about NPS overhead, it's not that significant. Adding up the NPS is pretty valid. -Tom
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.