Author: Dann Corbit
Date: 10:29:38 04/28/99
Go up one level in this thread
On April 27, 1999 at 20:03:51, James Robertson wrote: >On April 27, 1999 at 18:08:21, Dann Corbit wrote: > >>On April 27, 1999 at 18:05:46, Dann Corbit wrote: >>[snip] >>"How important will multithreading and multiple CPU's be for the future of chess >>and what will be the measurable impact on performance?" >>>> >>Since CCC asked for possible answers: >>[1] This year, chess programs that use multiple CPU's effectively will dominate >>those programs that do not. >>[2] Within two years, chess programs that use multiple CPU's effectively will >>dominate those programs that do not. >>[3] Within 5 years, chess programs that use multiple CPU's effectively will >>dominate those programs that do not. >>[4] Multiple CPU's will never be important in having the fastest chess program. >><< > >A good question and good answer suggestions, but I think [4] should be changed. >Do you mean fastest because the engine is fast or because the computer is fast? Actually, it could be either. I just meant is it possible that alternative technology (programs or hardware) will render multiple CPU's unimportant for increased performance. It seems to me that with .1 micron traces, they could put a large number of advanced CPU's on a single chip. Maybe we even get a full wafer slice with 100 CPU's on it. If there was a yield problem, you could tag some of them as non-functional. Can you imagine 100 K7's or 21264's? Vroom!!!
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.