Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: EGTB's and Fruit 2.2

Author: Uri Blass

Date: 01:16:29 11/03/05

Go up one level in this thread


On November 03, 2005 at 03:25:45, Mark Mason wrote:

>I have noticed that since upgrading Fruit 2.2 to Fruit 2.2.1, my average kn/s
>when playing engine to engine games has gone down from about 450 kn/s to 300kn/s
>(I have slow hardware !).
>
>Is this kind of slow down what one would expect as the result of accessing my
>EGTBs, and if so I would appreciate any advice as to whether people think this
>slow down is outweighed by the benefits of EGTBs - for blitz engine to engine
>games. Or is the net performance better without acessing EGTBs?
>
>Thanks,
>
>Mark

In most positions there is no tablebases hits so you are not going to see drop
in the nodes per seconds.

In the following position I get
[D]k7/ppp5/8/8/8/8/PPP5/K7 w - - 0 1

Fruit2.2 70770 knodes/40 seconds
Fruit2.2.1 68454 knodes/40 seconds

In the following position I get
[D]k7/pp6/8/8/8/8/PP6/K7 w - - 0 1

Fruit2.2 68560 nodes/40 seconds
48360 nodes/26 second
Fruit2.2.1 stop searching after 19020 knodes and 26 seconds and 298799 tablebase
hits.


You can see that in the second position there is a significant slow down and
Fruit2.2 is more than twice faster in nodes per second but on the other hand
Fruit2.2 only finished iteration 34 in the second position when Fruit2.2.1
finished iteration 63 at the same time

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.