Author: Vincent Diepeveen
Date: 17:29:15 02/26/99
Go up one level in this thread
On February 24, 1999 at 19:14:56, Richard A. Fowell (fowell@netcom.com) wrote:
>On February 24, 1999 at 16:58:15, Don Beal wrote:
>
>>On February 24, 1999 at 12:44:41, Vincent Diepeveen wrote:
>>
>>>My program runs at a 4 processor Xeon from Bob, but it is running
>>>at a PC-system, and it also available at 1 processor.
>>>
>>>Does that count too?
>>
>>You have to choose whether to enter the 4-processor version, or the
>>1-processor version. If you enter the 4-processor version it won't
>>be eligible for the current microcomputer title.
>>
>>This raises an interesting question, though. Would the participants
>>welcome another title category for multi-processors of the Xeon type?
>>They look like becoming widespread machines. The disadvantage of
>>too many titles is that it creates confusion and might be thought to
>>devalue the existing title. I'd be interested to hear the opinions
>>of the participants.
>>
>>Don Beal.
>
>For what it's worth, I don't think the time for "commercially available
>multiprocessor system" title is here, yet.
>
>How many multiprocessor entries do you have so far?
>How many entries could go either way? Crafty and Diep, it seems ...
Fritz is claiming to be parallel, so i expect them to be parallel, as
they had big capital letters that fritz played against anand using
a parallel machine and a parallel version of it.
To name a few:
Diep (currently still some bugs in, but i expect them to have
solved far for world champs start. i run at bob's quad
xeon under linux. So anyone with faster hardware may
send me an email and receive diep for free at least if i can
use that hardware at world champs. just need to use it
via internet during 8 days, and the type of system i don't
care, shared or non-shared. I'll get diep to work at it.
about tactics of diep at such a system, be warned!)
Crafty (please join bob!)
Fritz (quest engine, or they must have played with crafty against
anand otherwise)
SOS, (Rudolf Huber, is still looking for parallel hardware, anyone?
he runs in linux
and currently he has a 4 processor old ultrasparc as fast
as a todays K6-3 450 or PIII-500, his program is tactically
awfully strong)
Cilkchess (128 processors ultrasparc or something, also evaluation heavily
improved, and if he uses nullmove everywhere he'll kick some butts
further Ryan is a smart one and we can expect a
very good book. Don and Ryan form a good team!)
Zugzwang (128 processor Cray, no doubt the fastest hardware that'll play
there, but yeah how's the program nowadays? Heiner of course
should be able to give some good tips to improve it, so i expect
a way better zugzwang, and if i would have that hardware i'd
search like 16 plies or something, just using nullmove,
so we can expect something similar from ZZ and sure must
not challenge its tactics by doing a move that is seemingly
tactical possible but very antipositional)
Conners (60 processors at least and all are PII-300s, and he's
tactically not bad either, although his evlauation tuning
is also great. He's a big candidate for the title of course
with a program that has so little weak chains, and so much work
on the search and professional way of tuning all parts of it.)
And all folks that try to enter using a crafty clone, but the
'so called programmers' of it will be
knocked down before they enter the tournament hall (why would that be,
we sure didn't knock down any of the clones in Paris?).
No bionics or voyagers in Paderborn!
>Richard A. Fowell (fowell@netcom.com)
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.