Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: ACM1994

Author: Robert Hyatt

Date: 11:44:01 09/19/05

Go up one level in this thread


On September 18, 2005 at 12:07:32, Uri Blass wrote:

>On September 18, 2005 at 12:00:09, Vincent Diepeveen wrote:
>
>>On September 18, 2005 at 10:45:19, Robert Hyatt wrote:
>>
>># Name 1 2 3 4 5 P BU SB G
>>1 Deep Thought II 2w1  5b0  7w1  3b1  5b1  4  13½ 11  5
>>2 Zarkov          1b0  6b1  4w1  5b=  3w1  3½ 15  9¾ 5
>>3 Star Socrates   10w1 7b1  5w1  1w0  2b0  3  12½ 5  5
>>4 Now             6w=  10b= 2b0  8w1  9b1  3  10½ 5½ 5
>>5 Mchess Pro      8b1  1w1  3b0  2w=  1w0  2½ 16½ 7¾ 5
>>6 Cray Blitz      4b=  2w0  9w1  7b0  10w1 2½ 11  4  5
>>7 Wchess 9w1      3w0  1b0  6w1  8b0       2  13½ 4½ 5
>>8 Evaluator       5w0  9b0  10w1 4b0  7w1  2  10  2½ 5
>>9 Innovation II   7b0  8w1  6b0  10b1 4w0  2  10  2½ 5
>>10 Spector        3b0  4w=  8b0  9w0  6b0  ½ 12½ 1½ 5
>>
>>Zarkov from those days had no problems beating your 3 million nps Cray Blitz.
>>Nor had Wchess problems beating your 3 million nps Cray Blitz.
>>
>>I remember how Wchess a few years ago was online at icc. Old Crafty had no
>>problems beating it.
>>
>>Let's find those games somewhere in the archives... ...and see how strong
>>Cray Blitz tactical was...
>>
>>Vincent
>
>The problem was bugs in cray blitz and the cray blitz that beated Crafty was
>significantly better.
>
>Bob did not prove that the cray blitz that beated Crafty was not better than
>deep thought(I believe that it was better but there is no point to argue about
>it).
>
>Uri


I will remind everyone, including Vincent (again) that the primary reason there
is no longer a "Cray Blitz" is the difficulty of obtaining test time, to avoid
these types of bugs.  If we were lucky, we got to play 4-5 games per year before
the annual ACM/WCCC events.  We could test on non-Cray machines, using the pure
FORTRAN version of the code, but the search depth was about 1/2 the depth we
could reach on the Cray, which stressed the code much more than a shallow search
on a vax.  Not to mention the inability to test the cray assembly language at
all except when running on a Cray.

I simply got tired of the hassle, and decided to move to a more available
platform, namely the PC-compatibles including non-PC workstations.  It was a
wise decision.  Vincent only has to look at the troubles he had on the big NUMA
box he used last year, and he apparently got more test time than we usually did.
 It just doesn't work...

So there is no guarantee that the version that beat crafty had fewer bugs.  It
might well have had more.  There were just not enough games to say.  After 1000
games on ICC, I can be pretty sure my program won't crash during a game.  But
after just 5 I can't be sure of anything other than I just played 5 games...




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.