Hydra Computer

  • Thread starter Thread starter Tantale
  • Start date Start date
Tantale said:
The computer Hydra defeated one of the best world players Michael Adams
by 5.5 to 0.5
See http://www.jmrw.com/Chess/Adams/base.htm

I am looking about precise informations about this computer

Thanks

How precise? They aren't giving away the design or anything but if you
just google for "hydra computer chess" (without the quotes of course)
you will find a FAQ that describes the hardware and software and the
reasoning behind Hydra. Things like it being a 64 XEON processor machine
running with only half its normal complement during the match seemed to
be interesting.
 
kony said:
Why? It's mostly a dog-and-pony show.
Any desktop PC should beat any chess player given the right
application, *same* source code & db for the chess.

I don't think you are too familiar with the problems of computer
chess. Even with fast equipment it is no small feat to produce a
quality player. From my woodpusher standpoint I looked at a couple
of the games in that reference, and I am impressed. Too bad the
games were not annotated. Even the display is impressive.
 
I don't think you are too familiar with the problems of computer
chess. Even with fast equipment it is no small feat to produce a
quality player. From my woodpusher standpoint I looked at a couple
of the games in that reference, and I am impressed. Too bad the
games were not annotated. Even the display is impressive.

Note that I wrote "same source code" though. In other
words, just using any good modern PC to run the same
program, same logic the supercomputer did. I'll concede
that it wouldn't be as fast, perhaps having to make a few
moves before time ran out rather than being able to play
through as many scenarios as the supercomputer, BUT we can't
be sure that would even make a difference in the end.
 
kony said:
Note that I wrote "same source code" though. In other
words, just using any good modern PC to run the same
program, same logic the supercomputer did. I'll concede
that it wouldn't be as fast, perhaps having to make a few
moves before time ran out rather than being able to play
through as many scenarios as the supercomputer, BUT we can't
be sure that would even make a difference in the end.

Hmmn. There is still not a computer that can regularly beat the best chess
players. There are more possible combinations of moves in a game of chess
than there are atoms in the universe - so I have heard. I find it hard to
get my head around this.

ss.
 
Hmmn. There is still not a computer that can regularly beat the best chess
players.

Perhaps, I haven't spent enough time to get stats on that.
Even if it were true, I'd argue that it's a matter of the
code, not the computer per se.

There are more possible combinations of moves in a game of chess
than there are atoms in the universe - so I have heard. I find it hard to
get my head around this.


Where'd you hear it besides that forum you linked? I
wouldn't be surprised if there are more atoms in the chess
set itself than the # of possible move combinations.
 
kony said:
Perhaps, I haven't spent enough time to get stats on that.
Even if it were true, I'd argue that it's a matter of the
code, not the computer per se.




Where'd you hear it besides that forum you linked? I
wouldn't be surprised if there are more atoms in the chess
set itself than the # of possible move combinations.

It's something that I've heard many times, way before Deep Blue was created.
http://en.wikipedia.org/wiki/Shannon_number

Search for "chess+atoms+universe" in Google and you'll see that this has
been discussed for years.

I am not saying that I understand all this, but I do know that the game of
chess seems to be a great challenge for computer power that is not likely to
be on our desks for at least a few years.

ss.
 
I am not saying that I understand all this, but I do know that the game of
chess seems to be a great challenge for computer power that is not likely to
be on our desks for at least a few years.


Huh?

I"m sure there were chess games back in the 486 days, let
alone today.
 
Back
Top