Author: Slater Wold
Date: 04:06:11 03/23/01
Go up one level in this thread
On March 22, 2001 at 22:41:55, K. Burcham wrote: >i see where the 760mp chipset motherboard and the dual amd processor is expected >to be released in september. the article says it will most likely be 1500 mhz. >if this is like my other amd overclocked processors, it should overclock to at >least 1650 mhz., with the proper cooling. so this might go something like this: > >1650 x .9 = 1485kns (single processor) >1650 x 1.75 = 2887kns (dual processor) >1650 x 1.80 = 2970kns (dual processor with ddr) > >with the ddr ram, this might be closer to the 2970kns. > > >if this happens i dont see how the single processor programs stand a chance >against the top dual processor programs at these speeds. even with the new >single processor programs such as, tiger ll(with patch), century 3, etc., it >seems the kns gap between the single processor programs and the dual processor >programs is widening. it seems like they should do like we do in racing. > >your case would have to go through a tech line. >the case would have to be opened and checked for single, dual or quad >processors. then from there, all programs would be in one of these "classes". >class A would be unlimited, would include from 4 processors up. >class B would be dual processors only. >class C would be single processors only. >with no limit on ram or processor mhz. >and i dont think the programmers should have to weigh in. > >just a racing thought, >xracer While I agree that computers are getting faster, especially now with an AMD duals coming out, that the desire for people to want x2 or x4 processor machines and programs will rise. However, I don't think this necessarily says that x2 and x4 processors will be king. Take for example, me, I have a P4 1.5Ghz with 256MB of RAM. I use Fritz 6e with 64MB of hash. I usually get around 1100K nps. My account on FICS has a record of +8 -14 =12 with vippcomputers. Anyone who does not know who vipp is, he is a 1001mhz x 2 PIII with 1GB of RAM, running Deep Fritz. (I suspect he uses 256MB hash.) I know for a fact, he gets x2 what I get on the Crafty bench; therefore it's a good estimate to say he probably gets around x2 of my NPS on Fritz too. Now 2 of his wins, came from when autoplayer screwed up, so therefore, it is actually +8 -12 =12. 2 more wins! That's it. And usually when I am on FICS, I use a 67,000 game opening book and set it on "normal". Therefore, perhaps if I used a tournament book, there would be less wins, and more draws. There was a test a few years back again Deep Junior and Junior. Deep Junior using exactly x2 of the hardware that Junior was using. I believe Deep Junior won by .5 and it was mostly draws. This was a test, just to prove that, the deeper you go, the less the search means. I too am an avid (street) racer, so I know where you're coming from. However, I have come to understand that in computer chess, horsepower does not equal hardware. In other words, you can justify that 10HP = .1 of a second in the 1/4 mile. However, I believe the rule in computer chess is x2 mhz = 50 elo or so. If a program is 2500 on an P4 1.5Ghz and you dual it to x 2 1.5Ghz, you're only getting 50 elo. Not a huge increase. Well, not like going from 300HP to 600HP anyway. Ed, Robert, and all the programmers will tell you, there have been little advances in computer chess programming in the last 10 years. The difference between Fritz on a 486 DX-2 66 vs a P4 1.5Ghz would be one I think anyone would be able to see. Just my 2 cents. Slate
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.