HIARCS 2.0 Chess Program
Bryant Fujimoto
chess program for $125. They said that this program won the 1993
World Microcomputer Championship. Does anyone have any
experience with this program? How good is it?
Thanks for any information
Bryant Fujimoto
fuji...@denali.chem.washington.edu
Roy Eassa
>I recently received an add from Bookup offering the HIARCS 2.0
>chess program for $125. They said that this program won the 1993
>World Microcomputer Championship. Does anyone have any
>experience with this program? How good is it?
Yes, I was a beta tester. It is very strong -- probably about the same
as MChess Pro overall or maybe a bit better. It has a solid but not
spectacular feature set and has a lot of good chess "knowledge". It is
also quite good tactically (though a bit behind Genius 2.0 in that
regard). Overall, it is a leading contender and recommended.
Roy Eassa
Hal Bogner
The version of Hiarcs that won the 1993 World Micro was running on a Sun
workstation. The de Koning program that finished second was running on a
150 MHz DEC Alpha. PC's were represented only when you got down to 3rd/4th
(at 6-3); Chess Genius 2 on a Pentium and Nimzo on a 486.
-hal
Martin Zentner
|> D'aprŠs Bryant Fujimoto to ALL re HIARCS 2.0 Chess Program:
|>
|> BF´ I recently received an add from Bookup offering the HIARCS 2.0
|> ł chess program for $125. They said that this program won the 1993
|> ł World Microcomputer Championship. Does anyone have any
|> ł experience with this program? How good is it?
|>
|> I have it, and I am delighted with it.
I have seen Hiarcs in Munich and it played on a powerful SPARCstation. So
you have to expect, that the program will be weaker, if it plays on your home
PC. Most of the other programs did play on 486/50, a few already had Pentiums
and "The King" was playing on a DEC Alpha PC with 150 MHz.
So have in mind, that the results are platform-dependant !
Martin
Lisa Powell
BF´ I recently received an add from Bookup offering the HIARCS 2.0
ł chess program for $125. They said that this program won the 1993
ł World Microcomputer Championship. Does anyone have any
ł experience with this program? How good is it?
I have it, and I am delighted with it.
---
ţ RoseReader 2.10á P004120 Entered at [1000 BBS]
Mark Uniacke
Hi,
In article <2eq3qh...@tom.rz.uni-passau.de>,
Martin Zentner <zen...@pike.uni-passau.de> wrote:
>
>I have seen Hiarcs in Munich and it played on a powerful SPARCstation. So
>you have to expect, that the program will be weaker, if it plays on your home
>PC. Most of the other programs did play on 486/50, a few already had Pentiums
>and "The King" was playing on a DEC Alpha PC with 150 MHz.
>
>So have in mind, that the results are platform-dependant !
>
>Martin
I would like to clarify a few things you mentioned about Munich. Firstly, the
hardware that was used by the serious contenders was not ordinary PC hardware.
Heres a list:
HIARCS was running on a Sun Sparc 10 (thanks SUN!).
The King was running on a 150Mhz Dec Alpha.
Mephisto Genius 2 was running on a Pentium.
Mephisto Gideon Pro was running on a Pentium.
MChess Pro was running on a Pentium.
Most of the other competitors had 486/66 machines lent by Acer, some had
special purpose hardware or over clocked 486's.
The difference between the Pentiums and the Sparc in speed is small. The Dec
Alpha was definitely the fastest machine in the tournament. So it was not a
case of the program running on the fastest hardware winning. Results are
improved by playing on faster hardware BUT are not dependant! - there is no
substitute for a strong program.
Remember, HIARCS 2 also won the 1993 Uniform Platform Tournament a clear 3.5
points ahead of MChess Pro where ALL PROGRAMS were running on the SAME HARDWARE.
FYI: I am the chess programmer of HIARCS.
Cheers,
Mark
ma...@bnr.co.uk
The opinions and comments expressed herein are my own and do not in anyway
represent those of BNR Europe or Northern Telecom.
--
The opinions and comments expressed herein are my own and do not in anyway
represent those of BNR Europe or Northern Telecom.
Mark Mittelstaedt
-> clear 3.5 points ahead of MChess Pro where ALL PROGRAMS were running
-> on the SAME HARDWARE
Can you comment on what makes your AI so much stronger than the MChess
Pro without giving away any trade secrets? Is it some type of better
search method, heurstic knowledge, evaluation function or what?
Roy Eassa
HIARCS 2.x is a great program, definitely worth owning. However, I
wouldn't say it's significantly stronger than M Chess Pro 3.5, despite
that tournament result. It seems to be a bit stronger in "positional
understanding" than any other program, and right behind Genius 2.0 and
Genius 1.0 in tactical strength. However, M Chess Pro is probably second
best in positional understanding (but plays a riskier game, which can
backfire against other computers), and only a bit slower tactically. A
buyer cannot really go wrong with either product. And both authors are
constantly improving their programs and provide good support. That's
always a good sign.
Israel Silverman
-> wouldn't say it's significantly stronger than M Chess Pro 3.5,
-> despite that tournament result. It seems to be a bit stronger in
-> "positional understanding" than any other program, and right behind
-> Genius 2.0 and Genius 1.0 in tactical strength. However, M Chess Pro
-> is probably second best in positional understanding (but plays a
-> riskier game, which can backfire against other computers), and only a
-> bit slower tactically. A buyer cannot really go wrong with either
-> product. And both authors are constantly improving their programs
-> and provide good support. That's always a good sign.
How do CG2 and MC 3.5 do against each other?
Roy Eassa
I haven't had either new version long enough to do oodles of
testing. However, early tests show that, in solving real-game
move-and-win type puzzles, CG 2.0 is the best ever. In "speed" games, CG
2.0 and CG 1.0 are a little ahead of the pack, with CM4000 next. MChess
3.5 and HIARCS 2.1 are also very strong, but maybe a bit behind in speed
chess. In slow games, the four are all in the same ballpark, with
Zarkov 2.6 and Fritz2 somewhat behind. I imagine that Zarkov 3.0 will
come near closing the gap, but I haven't received my copy yet. I haven't
tested Gideon Pro at all, but it too is likely to be strong. Finally,
the ChessMachine 15 MHz programs Gideon 3.0 and King 2.0 are both very
strong but the CPU is about twice as slow as a good '486; if you give
them twice the time against the '486 programs, they score close to 50%.
Roy
Roy Eassa
> Thanks. I have MC 3.12, but haven't gotten 3.5 yet. I've heard
> that 3.5 does slower than 3.12 in problem sets. This suggests to
> me a slower tactical algorithm. Any thoughts?
In my tests, MC 3.5 is faster on some problems than 3.12, and
slower on others. Its results in speed games against other computers are
much better than before. (Oddly, Genius 2's winning percentage against
other computers is a lot lower than Genius 1's was.)