Collector's Corner..On Hiarcs In A Dedicated Chess Computer

Cubeman · Post by **Cubeman** » Mon Jul 04, 2011 1:55 am

Thanks, I will be first to purchase if it arrives.

Post by **Steve B** » Mon Jul 04, 2011 7:59 am

Mark Uniacke wrote: It looks so far like Revelation HIARCS is strongly challenging for the crown of world's strongest dedicated chess computer.

Which for me would make it the Strongest Chess Playing Entity in the World

Dedicated Chess Computers Rule Regards
Steve

Post by **Steve B** » Mon Jul 04, 2011 8:03 am

Game 8 ..R has white and opens with 1.Nc3

R is out of book after move 5 scoring its position -.04
H is out of book after R's 6th move and scores the position +.09
H's handling of its N's in this game was very entertaining for me to watch
Two full scale N re-deployments moving backwards to the back rank (ala
Petrosian )totally unanticipated by R resulting in two connected and passed P's on R's 3rd rank
H tops this off with an Exchange sacrifice.. and then..
a full R sacrifice(while facing a mate in one by R)
bringing the house down with the spectators in the gallery(me and the cat) showering the board with gold coins
[Round "8"]
[Date "2011.7.4"]
[White "Res II-Rybka 2.2n8"]
[Black "Rev-Hiarcs 13.3"]
[Time Control "1 Min Per Move"]
[Result "0-1"]

1.Nc3 Nf6 2.d4 d5 3.Nf3 c5 4.dxc5 Nc6 5.e3 e6 6.Bd3 Bxc5
7.O-O e5 8.e4 d4 9.Na4 Be7 10.c3 Bg4 11.h3 Bxf3 12.Qxf3
Qa5 13.Bc2 b5 14.Nc5 Bxc5 15.b4 Qb6 16.bxc5 Qxc5 17.Bd2 b4 18.c4 O-O
19.a3 a5 20.Ba4 Rfc8 21.Rfc1 Ne7 22.Ra2 Ng6 23.g3 Rcb8 24.h4 Nf8 25.Rca1
Qxc4 26.Rc2 Qa6 27.Rc5 Ne6 28.Rxe5 d3 29.axb4 axb4 30.Kg2 Nd4 31.Qf4 Ne2
32.Qf3 Nc3 33.Rg5 Kh8 34.Bxc3 bxc3 35.e5 Ng8 36.Qxf7 Qb7+ 37.Qxb7 Rxb7
38.Rg4 Rxa4 39.Rgxa4 c2 40.Rc4 Rb1 41.Ra8 d2 42.Rcc8 h6 43.Rxg8+ Kh7
44.h5 Rg1+ 45.Kh2 Rh1+ 46.Kxh1 d1=Q+ 47.Kg2 Qd5+

FINAL POSITION
[fen]R5R1/6pk/7p/3qP2P/8/6P1/2p2PK1/8 w - - 0 48[/fen]

R Resigns showing -9.99

The Hiarcs Juggernaut continues leading with..
6.5-1.5 Regards
Steve

Post by **Steve B** » Mon Jul 04, 2011 4:19 pm

Game 9 ..H has white for the last time in the match and opens with 1.Nc3
H was out of book after move 5 scoring the position -.07
H's 6th move was in R's book
R was out of book after H's 7th move scoring the position -.11 while H
scores it +-.00
H wins a P at move 30 and then goes on to outplay R in a R+P ending forcing R's resignation in 41 moves making it the shortest game in the match

[Round "9"]
[Date "2011.7.4"]
[White "Rev-Hiarcs 13.3"]
[Black "Res II-Rybka 2.2n8"]
[Time Control "1 Min. Per Move"]
[Result "1-0"]

1.Nc3 c5 2.e4 Nc6 3.Bb5 Nd4 4.Bc4 e6 5.Nge2 Nf6 6.O-O a6 7.Bb3 Qa5 8.d3
Nxb3 9.cxb3 Be7 10.d4 O-O 11.Qe1 Qc7 12.Bf4 d6 13.dxc5 Qxc5 14.Qd2 Bd7
15.Rfd1 Ng4 16.Ng3 f6 17.Rac1 g5 18.Nd5 Qxf2+ 19.Qxf2 Nxf2 20.Nxe7+ Kf7
21.Kxf2 gxf4 22.Ngf5 exf5 23.Nxf5 Bxf5 24.exf5 Rfc8 25.Rc3 Ke7 26.Kf3
Rxc3+ 27.bxc3 Rc8 28.Rd3 Rg8 29.Rd4 Kd7 30.Rxf4 Rc8 31.Rc4 Rxc4 32.bxc4
Kc6 33.Kg4 b5 34.cxb5+ axb5 35.a3 Kd5 36.Kh5 Ke5 37.g4 Kf4 38.h3 Ke4
39.Kh6 d5 40.Kg7 Kd3 41.g5

FINAL POSITION
[fen]8/6Kp/5p2/1p1p1PP1/8/P1Pk3P/8/8 w - - 0 41[/fen]

Hiarcs Leads 7.5-1.5

Blowout Regards
Steve

Post by **Steve B** » Mon Jul 04, 2011 8:52 pm

Game 10 ..R has white for the final game of the match and opens with 1.Nf3
R was out of book after move 9 scoring the position +.06
H was out of book after R's 10th move scoring the position -.26 while R scores it +.07
the game features an early tactical melee with R exchanging its Q for R+N+P
as the game progressed H won back one P,then another P and then a 3rd and finally the game to conclude the match

[Round "10"]
[Date "2011.7.4"]
[White "Res II-Rybka 2.2n8"]
[Black "Rev-Hiarcs 13.3"]
[Time Control"1 Min.Per Move"]
[Result "0-1"]

1.Nf3 Nf6 2.c4 g6 3.g3 Bg7 4.Bg2 O-O 5.O-O d6 6.Nc3 Nc6 7.Rb1 e5 8.d3 a5
9.a3 Bg4 10.Be3 Qc8 11.Qb3 Qd7 12.Qb5 Rfb8 13.Rfd1 a4 14.Nxa4 Ra5
15.Nxe5 Qe6 16.Nxc6 Rxb5 17.Nd4 Qe8 18.Nxb5 Bxe2 19.Rd2
Bg4 20.Re1 Qd7 21.f3 Be6 22.Nac3 d5 23.b3 Re8 24.Ree2 c6 25.Nd4 Qd6
26.b4 Nd7 27.Bf2 Ra8 28.c5 Qe7 29.a4 Bxd4 30.Bxd4 b6 31.Rc2 bxc5 32.bxc5
Ra5 33.Rb2 Qf8 34.Bf1 Nxc5 35.Rb6 Qc8 36.Reb2 Nd7 37.Rb7 Qd8 38.Bf2 Qf6
39.d4 Qxf3 40.R2b3 Qf5 41.Rb2 Nf6 42.Be1 Ng4 43.Re2 Qf3 44.Bd2 Bf5
45.Rb4 Bd3 46.Bg2 Qf5 47.Re8+ Kg7 48.Be1 Qd7 49.Rbb8 Qa750.Rg8+ Kh6
51.Rb4 Bc4 52.Kh1 Qxd4 53.Rb2 g5 54.Re8 Ra6 55.Re7 Ne5
56.Rf2 Kg6 57.Rd2 Nd3 58.Rxd3 Qxd3 59.Re8 c5 60.Ne2 Ra8 61.Bf1 Qf5
62.Bh3 Qxh3

FINAL Position Of The Match
[fen]r3R3/5p1p/6k1/2pp2p1/P1b5/6Pq/4N2P/4B2K w - - 0 63[/fen]
Its Mate in 8

Hiarcs wins the Match 8.5-1.5
Using Rev-Rybka's SSDF rating of 2632 as a base.. Hiarcs achieves a Performance rating for this match of 2912!

Oh My Regards
Steve

P.S.Spacious Mind is currently conducting a 12 game match(1 Min/Avg.) Rev-Hiarcs 13.3 Vs.Rev-Shredder 12(SSDF ELO 2725) in this thread:
http://hiarcs.net/forums/viewtopic.php?t=4269
the match stands at 1-0 Hiarcs(as i write)

SirDave · Post by **SirDave** » Tue Jul 05, 2011 5:29 pm

Steve, thanks again for taking the time to run this match. It's unfortunate that the informal, unpublicized nature of it meant that relatively few in the chess community are/were aware of it. To me, for the dedicated chess board world, it was analogous, in the human grandmaster world, to someone scheduling an impromptu match between Kramnik & Anand just for the benefit of a few people on a forum.

Again, given recent history, the end result seems fitting. Heaven forbid it had turned out the opposite way. Congratulations to Mark Uniacke.

Of course, we still have Nick's Hiarcs/Shredder match.

Nail-biting regards,
Dave

Post by **Steve B** » Tue Jul 05, 2011 9:44 pm

Thanks Dave
i am glad you found the match interesting
i am looking forward to the match Vs. Shredder 12
I'm thinking we are looking at a Hiarcs 13.3 rating significantly higher then the estimated 2750
Time Will Tell Regards
Steve

Post by **Mark Uniacke** » Fri Jul 08, 2011 1:40 am

Thanks for playing this interesting match Steve. I nice start for Hiarcs on its first dedicated chess computer!

It is also particularly interesting because of the varied openings and the way that Hiarcs dominated Rybka both positionally and tactically in all game phases. I notice in no game was the opening book a deciding factor and the manner of victory shows just how much stronger Hiarcs is than Rybka on the Revelation.

I am hoping Hiarcs can break through the 2800 Elo level on the Revelation module.

Way back in late 1993 Saitek had the opportunity to have Hiarcs on their Sparc module, at the time Hiarcs had just won the World Microcomputer Championship in Munich on Sparc processor and ran at about 2K nps on the Sparc/20Mhz CPU used in their module so would of made a very strong program back then for a dedicated chess computer. Unfortunately, Saitek decided not to have Hiarcs on their module, a great pity since Hiarcs ran very nicely on the Sparc.

ricard60 · Post by **ricard60** » Fri Jul 08, 2011 1:54 am

Mark Uniacke wrote:Thanks for playing this interesting match Steve. I nice start for Hiarcs on its first dedicated chess computer!

It is also particularly interesting because of the varied openings and the way that Hiarcs dominated Rybka both positionally and tactically in all game phases. I notice in no game was the opening book a deciding factor and the manner of victory shows just how much stronger Hiarcs is than Rybka on the Revelation.

I am hoping Hiarcs can break through the 2800 Elo level on the Revelation module.

Way back in late 1993 Saitek had the opportunity to have Hiarcs on their Sparc module, at the time Hiarcs had just won the World Microcomputer Championship in Munich on Sparc processor and ran at about 2K nps on the Sparc/20Mhz CPU used in their module so would of made a very strong program back then for a dedicated chess computer. Unfortunately, Saitek decided not to have Hiarcs on their module, a great pity since Hiarcs ran very nicely on the Sparc.

Hi Mark,
Did not know that you test hiarcs on Sparc module. Did you ever made any test games with Sparc/Hiarcs against other machines?

secret games regards
Ricardo

Post by **Mark Uniacke** » Fri Jul 08, 2011 2:08 am

Hiarcs played on a Sparc in Vancouver WMCC 1991, Madrid WCC 1992, London Olympiad 1993 and Munich WMCC 1993.

It ran on a Sparc 1/25Mhz at about the same speed as a 486/50-66Mhz. Faster Sparcs were used in 92 & 93, I think a Sparc 2/40Mhz.

So it played against many well known computers at the time.

Interestingly in Munich it played the King which was running on a 150Mhz Dec Alpha, now that was like a R30 on steroids!

Post by **Steve B** » Fri Jul 08, 2011 8:00 am

Mark Uniacke wrote:Hiarcs played on a Sparc in Vancouver WMCC 1991, Madrid WCC 1992, London Olympiad 1993 and Munich WMCC 1993.

It ran on a Sparc 1/25Mhz at about the same speed as a 486/50-66Mhz. Faster Sparcs were used in 92 & 93, I think a Sparc 2/40Mhz.

So it played against many well known computers at the time.

Interestingly in Munich it played the King which was running on a 150Mhz Dec Alpha, now that was like a R30 on steroids!

Hi Mark
great stuff reading your posts
each one reads like another chapter in computer chess history

you know..
i have two Sparc modules ..
the normal 1 MB module and the 4MB module
the 4MB competed in Alains Paris Tournament
http://www.zanchetta.net/echecs/Paris2008/

it came in second to last place
now can you imagine how Sparky would have finsihed if it had Hiarcs loaded into it?

I'm Thinking ..

5-Zip ..You Cant Touch This Regards
Steve

Post by **Mark Uniacke** » Wed Jul 13, 2011 12:14 am

At the time I was discussing Hiarcs 2.1 with Saitek, Hiarcs 2.1 was the 93 World Champion but was untested by the SSDF, however we have Hiarcs 2.0 as a reference point.

From my knowledge of Hiarcs' performance on the Sparc CPU and the improvement in v2.1 I can make a very good estimate for a Sparc 20Mhz/1Mb. I think based on the current SSDF rating list levels it would of rated at about 2168 Elo, just over 40 elo above the Spracklen Sparc module.

Also its worth considering if Saitek had released a SuperSparc 50Mhz model as they used in Madrid and Munich then it would of been very interesting. The Sparc CPU was great for chess with lots of sliding registers and a decent cache.

Post by **Steve B** » Wed Jul 13, 2011 12:24 am

Mark Uniacke wrote: The Sparc CPU was great for chess with lots of sliding registers and a decent cache.

and yet Saitek was the only Company to use the Sparc and only in one dedicated computer

What A Crying Shame Regards
Steve
P.S.. one gripe i have with it are the cooling fans..it is a very noisy computer

Post by **Mark Uniacke** » Wed Jul 13, 2011 12:35 am

Yes a pity, but I think it was the cost.

I know Saitek spent a fortune on the program and module development and they had expected to win a world title somewhere in the 91-93 years. I don't think they had the will to continue with it after that and that is why they could only offer so little for Hiarcs at that time.

On reflection I should of taken the job on as part of a loss leader to greater things, but at the time I had a very full time job as a software engineer for STC/BNR/Nortel as well as lots of development on Hiarcs PC and I could not justify doing the Sparc work almost for free, but maybe it would of been a good idea long term.

Missed opportunity? regards,
Mark

Fluppio · Post by **Fluppio** » Wed Jul 13, 2011 8:58 am

Hi Mark,

missed opportunity? Maybe not

Do you think it's possible to create a "Hiarcs 2.1 SPARC"-Version nowadays? Don't know how far you had gone in the business with Saitek at that time, but this would be a very interesting project.

Best regards
Peter