Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: APHID , advances in ICCA #8

Author: Robert Hyatt

Date: 19:13:07 09/18/01

Go up one level in this thread


On September 18, 2001 at 17:19:31, Gian-Carlo Pascutto wrote:

>On September 18, 2001 at 16:56:45, Robert Hyatt wrote:
>
>>On September 18, 2001 at 14:49:58, Vincent Diepeveen wrote:
>>
>>>On September 18, 2001 at 12:42:24, Gian-Carlo Pascutto wrote:
>>>
>>>>On September 18, 2001 at 11:54:45, Vincent Diepeveen wrote:
>>>>
>>>>>All testresults they published are based upon 5 to 8 ply searches
>>>>>and based upon a program called TheTurk. No mention of crafty speedups,
>>>>>and if they are then it's still 5 to 8 ply.
>>>>
>>>>Literal quote:
>>>>
>>>>'..and accomplishes 11 and 12 ply searches in approximately 260
>>>>seconds using 32 processors'
>>>
>>>Advances in computerchess 8 only TheTurk is quoted there
>>>with only 5 to 8 ply searches which are put in graphs.
>>>
>>>Your weird conclusions there is not a word on that in AICC8.
>>>
>>>>Given that they used 230 seconds on 16 processors for the speedup
>>>>data, it's based on 10 and 11 ply searches.
>>>
>>>that's a nonsense conclusion. There is ONLY a graph from 5 to 8 ply
>>>searches.
>>>
>>>>The graphs in the technical report show increasing APHID performance
>>>>as depth gets larger, and not the other way around.
>>>
>>>Of course it goes from 5 to 8 ply here!
>>>
>>>>--
>>>>GCP
>>
>>
>>I think you are looking at two different papers.
>
>I'm quoting the numbers directly out of Brockingtons thesis and
>the technical reports up at UAB.
>
>--

>GCP


That is my point.  Vincent is using a very old Advances in Computer Chess
book with an article with a similar name.  Which is why you are quoting one
set of numbers, he is quoting another.  You are displaced in time 3-5 years
from him...



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.