Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: hash table size - is a power of 2 still an advantage these days?

Author: Uri Blass

Date: 14:37:06 09/24/03

Go up one level in this thread


On September 24, 2003 at 16:38:16, Sune Fischer wrote:

>On September 24, 2003 at 15:49:15, Uri Blass wrote:
>
>>Did you compare speeds to find out that it is 0.05%?
>>I suspect that it is clearly bigger than it.
>
>The modulo is a binary operator, I would not expect it to take hundreds of
>clocks.
>It's about the same speed as a division IIRC, (you don't need the full fraction,
>only the remainder).
>
>But more importantly, it's a design decision to limit the programs capabilites
>in this way. I prefer to let the user (incl. myself) make use of *all* the
>memory, afterall I didn't put 512 MB of memory in my machine to run with a mere
>256 MB of hash! :)
>
>Of course it matters most when doing long analysis, not very important for blitz
>games.

How much it matters?

I read claims in the past that doubling the hash only gives 6-7% speed
improvement and if you lose 1-2% from division then it may be important.

The only way to know is testing and from my experience 1-2% speed reduction
because of division is possible.

I also know from experience that when I use big hash tables the program may be
very slow in the first minutes so I prefer not using more than half of the RAM
of the machine for hash tables for long analysis.

I want to be able to do other things during analysis without having big
problems.

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.