Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: General efficiency question

Author: Daniel Clausen

Date: 01:37:01 04/22/02

Go up one level in this thread


Hi

On April 22, 2002 at 04:03:46, Pham Minh Tri wrote:

>On April 22, 2002 at 03:46:08, Daniel Clausen wrote:
>
>>Whenever I can choose between
>>
>>(a) write my programs in a clean/readable way
>>(b) optimize to get the last 5% in speed
>
>Oops, I think the proportion should be from 50-500%, up to the gap between
>skills when someone starts and when he becomes a computer chess expert ;)

These things surely happen, but they're not the rule, IMO. Usually it's a change
in the program logic which results to these big speed improvements, whereas the
small so-called improvements are not even guaranteed to be faster the next
day...

If you didn't profile your chess engine (I think you're working on one too,
AFAIK) I suggest you invest some minutes first and make a little list of the
functions you think are the most time-critical. Then profile your engine and
compare the results.. Expect to be surprised. :) [Ok, if you put everything in
main(), then it's easy, but that doesn't count! :)]

Sargon



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.