Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: How important is a big hash table? Measurements...

Author: Robert Hyatt

Date: 18:49:34 03/30/03

Go up one level in this thread


On March 30, 2003 at 20:36:12, Jeremiah Penery wrote:

>On March 29, 2003 at 17:26:35, Robert Hyatt wrote:
>
>>I don't see anything "contrary to Hyatt's assertion."
>>
>>It _might_ depend on implementation details as well.  I'll try to run a
>>similar test tonight and post the results.  I haven't tested it since
>>"Komputer Korner" posed the question and I tried it a few years ago in
>>r.g.c.c
>
>And here is the data you came up with then:
>
>depth 9:
>time     bytes   entries     nodes searched
>-------------------------------------------
> 1:05     96kb        6K         12,139,882
> 1:00    192kb       12K         11,245,725
> 54.5    384kb       24K          9,995,634
> 55.1    768kb       49K         10,131,840
> 52.3     1.5M       98K          9,567,818
> 46.3       3M      196K          8,463,339
> 44.5       6M      392K          8,119,062
> 43.5      12M      800K          7,932,811
> 43.0      24M      1.5M          7,833,488
> 42.9      48M      3.0M          7,802,585
> 42.7      96M      6.0M          7,779,999
> 42.7     192M     12.0M          7,779,122
> 42.7     384M     24.0M          7,778,994
>
>depth 11:
>time     bytes   entries     nodes searched
>-------------------------------------------
>13:13     96kb        6K        157,085,451
>12:03    192kb       12K        142,633,162
>10:31    384kb       24K        123,762,238
> 9:28    768kb       49K        110,838,220
> 8:38     1.5M       98K        100,802,339
> 7:48       3M      196K         90,979,000
> 7:22       6M      392K         85,975,960
> 6:53      12M      800K         80,347,212
> 6:32      24M      1.5M         76,465,119
> 6:22      48M      3.0M         74,738,532
> 6:13      96M      6.0M         73,253,374
> 6:05     192M     12.0M         71,581,397
> 6:03     384M     24.0M         71,156,722
>
>So, varying hash size from 3MB to 384MB only yields about 25% decrease in nodes
>searched at depth 11, and less than 9% decrease at depth 9.


Sure.  But _now_ a normal search depth is 12-13 plies.  Which changes things a
lot.  I posted a set of new data for the current Crafty, and the results seem
to follow a similar trend, except that the optimal size has increased quite a
bit due to faster hardware...




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.