Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Difference Statistically: Hardware - Shredder vs. Hydra?

Author: Jonas Bylund

Date: 05:12:03 08/19/04

Go up one level in this thread


>Since the "Nimzo" program has always been lagging behind the top programs I
>would assume the hardware Hydra is on is vastly superior to the Quad Shredder is
>on.  Am I wrong?
>Jim

Well by default i think you are wrong at one point, Just because of the sheer
nature of hardware programming, you can almost be 100% certain that Hydra has
very little to do with Nimzo X in terms of how it is put together, also the
amount of knowledge would be much greater in Hydra than in Nimzo X since there
is virtually no slowdown kn/s wise.

There is no telling wether the "vast superior" hardware of Hydra is a benefit
over a Quad Opteron, maybe the programming is just better, we don't know. In
other words, at this stage Hydra might need 10x the processing power to equal
the benefit that Shredder get's from the Quad Opteron in terms of strength vs.
potential. Let's say for example that on one processor (2.2 Ghz Opteron)
Shredder is at 65% of it's true potential and on a Quad it is at 80%, whereas
Hydra ,because we don't really know much about it, running on one card could be
at say 10% of it's potential and on the hardware against Shredder it might be
only at 20%.



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.