Author: Ed Schröder
Date: 04:22:08 06/21/00
Go up one level in this thread
On June 21, 2000 at 04:58:36, blass uri wrote: >On June 21, 2000 at 03:17:10, Ed Schröder wrote: ><snipped> >>I once rewrote a time consuming part of EVAL storing all possible results >>in a pre-compiled table. 2-3 table lookups replaced a large piece of code. >>I estimated the gain in speed on about 12-14%. In practice it was only 1% >>faster. Reason was blowing up the data cache. > >I guess that the change in speed is dependent on the computer and it is possible >that the same change that give you only 1% will give in another computer more >than 1% and will do you slower in another computer. Yes. >The question is what we can expect about the future. > >Can we expect this to be more than 1% in tommorow hardware? Yes, as data-cache will become bigger and faster. But I mainly wrote the above to explain that faster code not always gives faster timings. There are more of such issues such as keeping the size of the chess engine in first level cache. Ed >Uri
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.