Amiga 68060 Sargon 3 is scary good!

This forum is for general discussions and questions, including Collectors Corner and anything to do with Computer chess.

Moderators: Harvey Williamson, Steve B, Watchman

Forum rules
This textbox is used to restore diagrams posted with the fen tag before the upgrade.
User avatar
spacious_mind
Senior Member
Posts: 3999
Joined: Wed Aug 01, 2007 10:20 pm
Location: Alabama
Contact:

Amiga 68060 Sargon 3 is scary good!

Post by spacious_mind »

Last night I had a test game between Amiga 68060 Sargon 3 against Virtua Chess playing on Qemu a dos emulator created by Franz (nickname Fhub).

Amiga 68060 Sargon 3

Image

Picture above is Sargon 3 as it looks on an Amiga computer. The above pic is from the WINUA emulator that I use when I play it on my computer. It is the same setting that I use for Sargon 3 in my Division 3 Swiss League Tournament. I play the emulator on a Pentium I7 so it plays quite fast, probably quite a bit faster than what a real Amiga 68060 would play at. Fastest as I far I know Amiga 68060 played at 50 MHz but there exist also quite a few accelerator cards so not sure how fast you can actually get an Amiga 68060 to perform on.

Image

Above is Virtua Chess playing on Qemu. I have been trying for ages to see if I can get Virtua Chess to work on DOSBox without success. There seems to be some Graphic Card incompatibly when you use DOSBox and I have not been able to get around the incompatibility. If anyone has figured out how to use Virtua Chess with DOSBox then please share how as I really wanted to include it in a future Division 2 or Division 1 of my tournament.

Anyway out of desperation I tried Virtua Chess out again with Qemu last night.

At https://chessprogramming.wikispaces.com/Virtua+Chess it states that Virtua Chess runs at 20,000 nodes per second on a Pentium 90. (This is mainly the reason why I am playing my DOSBox programs at 69,300 cycles trying to emulate as close as I can a Pentium 90 as there are many P90 references in magazines in the past and not least being the old SSDF rating lists that tested on a P90).

The computer I used last night for Qemu Virtua Chess runs at a speed of 4.2 GHz with 8 cores. It played in a pretest 2.625 million nodes in 29 seconds. Dividing these nodes by the 29 seconds you come to an approximate speed of a Pentium 90 running at an overdrive speed of 400 MHz.

If you read https://chessprogramming.wikispaces.com/Virtua+Chess you will see that Virtua Chess beat David Bronstein in an Aegon Tournament in 1995 and Cumulus 2 which was the programmers (Marc-Francois Baudot and Jean-Christophe Weil) non commercial version also finished tied 2nd in the 1992 7th World Chess Computer Championship in Madrid.

Therefore knowing that Amiga 60060 Sargon 3 was playing a strong opponent running at a Pentium 90 overdrive speed of 400 MHz, I really did not expect much success for Sargon 3 but played it anyway as I thought it might be a useful enough test to see if I can include Qemu Virtua Chess in my Swiss Tournament. Therefore I used Amiga 68060 as a gage.

Test Match


[Event "Computer Test Match"]
[Site "Alabama"]
[Date "2017.12.07"]
[Round "1"]
[White "AM68060 Sargon 3, 30S."]
[Black "QEMU Virtua Chess, 30S."]
[Result "1-0"]
[ECO "C29"]
[WhiteElo "2400"]
[BlackElo "2400"]
[PlyCount "225"]
[EventDate "2017.12.07"]
[EventCountry "USA"]

1. e4 e5 2. Nc3 Nf6 3. f4 d5 4. fxe5 Nxe4 5. Nf3 Bg4 6. Qe2 Nxc3 7. dxc3 Nc6 8. Bf4 Be7 9. Qe3 O-O 10. O-O-O b6 11. Qd2 Be6 12. Bd3 Qc8 13. Qf2 Bf5 14. Qg3 Bxd3 15. Rxd3 Qe6 16. Rhd1 Rfd8 17. Ng5 Bxg5 18. Qxg5 Ne7 19. c4 c6 20. Kb1 a5 21. Bg3 a4 22. Bf2 h6 23. Qg3 Ra6 24. cxd5 cxd5 25. Rf3 Ng6 26. Bd4 Re8 27. Rdf1 Ra7 28. h4 Kh7 29. Rf5 a3 30. h5 Nh8

[fen]4r2n/r4ppk/1p2q2p/3pPR1P/3B4/p5Q1/PPP3P1/1K3R2 w - - 0 31[/fen]

The above is the position after the move 30. ... Nh8. By now I was becoming quite excited about Sargon 3's play as it had built up a strong position and a good lead. However in the above position Sargon 3 made a mistake with 31. bxa3. Much better was 31. R5f4! or 31. b3.

After 31. bxa3 Qemu Virtua Chess 400 MHz got back into the game.

31. bxa3 Rea8 32. Bb2 Ra4 33. R5f4 Re4 34. Rd1 Raa4 35. Rxe4 Rxe4 36. Qd3 Qf5 37. Qxd5 Rh4 38. Qd3 Rxh5 39. g4 Qxd3 40. Rxd3 Rg5 41. Bd4 Ng6 42. Re3 Rxg4 43. Bxb6 Nf4 44. Rf3 Kg6 45. Kb2 Kf5 46. a4 Ke6 47. a5 Nd5 48. c3 Rc4 49. Kc2 h5 50. a3 Ra4 51. Rg3 g6 52. Bd8 Rc4 53. Kb3 Rc8 54. Bg5 Kxe5 55. c4

[fen]2r5/5p2/6p1/P2nk1Bp/2P5/PK4R1/8/8 w - - 0 55[/fen]

After 55 moves after Sargon 3 played 55. c4 the position is better for Virtua Chess. I really expected at this stage that the 400 Mhz is too much for Sargon 3 to cope with, especially in an endgame. However in the above position Virtua Chess may have missed its best continuation which was 55. ... Nf6. Playing 55. ... Nc7 allowed Sargon 3 to equalize again.

Nc7 56. Re3+ Kd4 57. Re7 Rb8+ 58. Ka2 Ne6 59. Bc1 Kxc4 60. Rxf7 Nc5 61. Rf6 Ne4 62. Rb6 Nc3+ 63. Ka1 Rf8 64. a6 Rf2 65. Rb4+ Kd3 66. Rb2 Rf5 67. Rd2+ Ke4 68. Rg2 Rf6 69. a7 Ra6 70. Rxg6 Rxa7 71. Rg3 Rc7 72. Bd2 Nd1 73. Rh3 Rc5 74. Kb1 Nf2 75. Re3+ Kd4 76. Rf3 Nd3 77. a4 Rc7 78. Rh3 Rb7+ 79. Ka1 Rh7 80. a5 Kc4 81. Ka2 Nf2 82. Rc3+ Kb5 83. Rb3+ Kc4 84. Rb4+ Kd3 85. Bg5 Ne4 86. Bd8 Rh6 87. Bb6

[fen]8/8/1B5r/P6p/1R2n3/3k4/K7/8 w - - 0 87[/fen]

87 moves into the game, I was just a couple of moves away from adjudicating the game as a draw and let Sargon 3 enjoy a draw against a strong and fast opponent. I had no expectations that Sargon 3 would get anything other than a draw but worried that it might lose. Losing was a concern because of the faster speed and also what I read at https://chessprogramming.wikispaces.com/Virtua+Chess where it states:

The program has perfect knowledge of KPK endgames

I figured that with Black's passed pawn on h5 that a blunder was more likely to come from Sargon 3 if it was not careful.

However, surprise, surprise in the above position Virtua Chess blundered by playing 87. ... Nd2? and Amiga 68060 Sargon 3 noticed it immediately and 88. a6!

After this Sargon 3 very skillfully maneuvered its pieces correctly to win this game. Amazing!!

Well this showed that at least with KNRBP versus KNRNP, Sargon 3 was more perfect in an endgame than Virtua Chess :)

Nd2 88. a6 Rc6 89. a7 Rc2+ 90. Ka3 Rc3+ 91. Ka4 Nc4 92. Rb5 Rc2 93. Ba5 Kd4 94. Rd5+ Ke4 95. Kb3 Nxa5+ 96. Kxc2 Kxd5 97. a8=Q+ Nc6 98. Qg8+ Kd4 99. Qg5 Nb4+ 100. Kd2 Nd5 101. Qxh5 Nf4 102. Qh8+ Ke4 103. Qf6 Nd5 104. Qe6+ Kd4 105. Qf5 Kc4 106. Qe5 Nb4 107. Qc3+ Kb5 108. Qd4 Nc6 109. Qd5+ Kb6 110. Qd6 Kb5 111. Kc3 Na5 112. Qb4+ Kc6 113. Qxa5 1-0

Final Position

[fen]8/8/2k5/Q7/8/2K5/8/8 w - - 0 113[/fen]

Lichess Evaluation

https://lichess.org/ljHca0Hy

AM68060 Sargon 3, 30S. (2400)
7 Inaccuracies
0 Mistakes
0 Blunders
17 Average centipawn loss

QEMU Virtua Chess, 30S. (2400)
5 Inaccuracies
5 Mistakes
2 Blunders
28 Average centipawn loss

I am continuously surprised at how well Amiga 68060 Sargon 3 plays, especially its endgame prowess. The Amiga version came out in 1988. I think that the Spracklen's having lost interest in chess never really realized how good this program was. With a little more attention and right hardware they may have had a World Champion between the years 1988 and 1994 and they did not know it!

Move over Richard Lang and Mephisto!! :)

Sargon 3 with new Amiga PC power is as relevant today as it might have been in 1988!

Scary Good!!!! ... regards

ps.... as for Virtua Chess even if I cannot get it to work with DOSBox, I will let it play in the Swiss Tournament using Qemu as an exception to the rule of Pentium 90.
Nick
User avatar
Bryan Whitby
Senior Member
Posts: 1001
Joined: Wed Feb 18, 2009 9:57 pm
Location: England

Post by Bryan Whitby »

Hi Nick

I was always under impression that Chessmaster 2100 series was a slightly updated version of Sargon 3 program. Maybe you could play a few games on the Amiga to see?

Regards
Bryan
User avatar
spacious_mind
Senior Member
Posts: 3999
Joined: Wed Aug 01, 2007 10:20 pm
Location: Alabama
Contact:

Post by spacious_mind »

Chessmaster Ireland wrote:Hi Nick

I was always under impression that Chessmaster 2100 series was a slightly updated version of Sargon 3 program. Maybe you could play a few games on the Amiga to see?

Regards
Bryan
Hi Bryan,

I have it is not as strong and also it is prone to bugs when it plays at 68060. I have pretty much tried every program that exists for the Amiga on the Emulator and it is clear to me that Sargon 3 is clearly better than all of them. It seems to be one of those rare old programs that actually gets better and better with more and more speed. I don't think with my I7 that I have reached that curve yet where its improvement slows down.

Maybe in a future I8, I9 or I10 we find that curve but its not there yet using the Amiga WINUAE emulator.

Best regards

ps.. you know the funny thing is as I play the games often I have SF8 running on 1 core on my older 2 core laptop (to not slow it down) while I multitask with it recording and following the game in Chessbase. Over the series of games that I have played in Division 3 and even yesterday on quite a few moves, I am amazed at how often Sargon 3 plays a move that was not initially one of the best two choices that SF8 shows in its analysis, but after having to re-analyze the Sargon 3 move, it then miraculously comes up with a top evaluation for that move :) My laptop only reaches about 26 ply in those 30 seconds between moves (so its not 40 ply). But then again Sargon 3 finds great moves at 9/10 ply :)

Perhaps it should be playing that Amazon computer thingy instead of Stockfish on equal terms? :P

Considering Sargon 3 search was created in 1988.... who knows perhaps it has a better method than what has been evolving over the past 20 years though engine development.

Maybe the engine guys have been barking up the wrong search tree all these years ... LMAO.... :shock:

Best regards
Nick
User avatar
Fernando
Admiral of the Fleet
Posts: 3059
Joined: Tue Jul 31, 2007 4:35 pm
Location: Santiago de Chile

Post by Fernando »

It is too much to ask you if you can send me a package with the emulator of amiga and sargon 3 ermbedded inside, ready to run in my window 10??
I have tried two or three times to get both things running, but it looks like my computational skill is not good enough.

Fern
Festina Lente
User avatar
spacious_mind
Senior Member
Posts: 3999
Joined: Wed Aug 01, 2007 10:20 pm
Location: Alabama
Contact:

Post by spacious_mind »

Fernando wrote:It is too much to ask you if you can send me a package with the emulator of amiga and sargon 3 ermbedded inside, ready to run in my window 10??
I have tried two or three times to get both things running, but it looks like my computational skill is not good enough.

Fern
Hi Fernando,

For you anything :) Will prepare it and send it to you.

Regards
Nick
Nick
User avatar
Fernando
Admiral of the Fleet
Posts: 3059
Joined: Tue Jul 31, 2007 4:35 pm
Location: Santiago de Chile

Post by Fernando »

spacious_mind wrote:
Fernando wrote:It is too much to ask you if you can send me a package with the emulator of amiga and sargon 3 ermbedded inside, ready to run in my window 10??
I have tried two or three times to get both things running, but it looks like my computational skill is not good enough.

Fern
Hi Fernando,

For you anything :) Will prepare it and send it to you.

Regards
Nick
You are a real friend ANd a chess prince, Nic.
As always, I keep myself indebted to you. Anything you could need from me -from my vast vaults of strange things in ches stuff- just ask...

Fern
Festina Lente
User avatar
spacious_mind
Senior Member
Posts: 3999
Joined: Wed Aug 01, 2007 10:20 pm
Location: Alabama
Contact:

Post by spacious_mind »

Fernando wrote:
spacious_mind wrote:
Fernando wrote:It is too much to ask you if you can send me a package with the emulator of amiga and sargon 3 ermbedded inside, ready to run in my window 10??
I have tried two or three times to get both things running, but it looks like my computational skill is not good enough.

Fern
Hi Fernando,

For you anything :) Will prepare it and send it to you.

Regards
Nick
You are a real friend ANd a chess prince, Nic.
As always, I keep myself indebted to you. Anything you could need from me -from my vast vaults of strange things in ches stuff- just ask...

Fern
Thanks and you are welcome. Let me know if it works for you this time :)

Regards
Nick
User avatar
Fernando
Admiral of the Fleet
Posts: 3059
Joined: Tue Jul 31, 2007 4:35 pm
Location: Santiago de Chile

Post by Fernando »

As I told in a mail, it does not.
Configuration does not offer nothing
Festina Lente
User avatar
Bryan Whitby
Senior Member
Posts: 1001
Joined: Wed Feb 18, 2009 9:57 pm
Location: England

Post by Bryan Whitby »

Wiki
Sargon III was a complete rewrite, using quiescence search, and a transposition table with BCH hashing.
Sargon III was ported to the 68000, to be the first third-party executable software for the Macintosh.
The 8086 Sargon III port for PC was done by Kevin Leavelle in 1984.
Three Sargon 3 software versions played the WMCCC 1984.
The Apple II came 10th with 4/7, the Macintosh 12th with 3½/7 and the 8088 Compaq PC 17th with 2½/7.

Interesting video
Sargon III playing on the Amiga.
https://youtu.be/ZeEBcDm1uKg


User avatar
spacious_mind
Senior Member
Posts: 3999
Joined: Wed Aug 01, 2007 10:20 pm
Location: Alabama
Contact:

Post by spacious_mind »

Chessmaster Ireland wrote:Wiki
Sargon III was a complete rewrite, using quiescence search, and a transposition table with BCH hashing.
Sargon III was ported to the 68000, to be the first third-party executable software for the Macintosh.
The 8086 Sargon III port for PC was done by Kevin Leavelle in 1984.
Three Sargon 3 software versions played the WMCCC 1984.
The Apple II came 10th with 4/7, the Macintosh 12th with 3½/7 and the 8088 Compaq PC 17th with 2½/7.

Interesting video
Sargon III playing on the Amiga.
https://youtu.be/ZeEBcDm1uKg


Yep, it makes sense that it was first on Apple since they were contracted to Apple from the 6502 days onwards through Hayden Software. Apple of course owned Macintosh.

You video link does not work it plays some other game video.

Thanks
Nick
Nick
User avatar
Bryan Whitby
Senior Member
Posts: 1001
Joined: Wed Feb 18, 2009 9:57 pm
Location: England

Post by Bryan Whitby »

Hi Nick

Very strange that.

I have just tested the video link on my phone and it takes me to the correct Amiga Sargon III video.

Maybe someone else can advise what video they can see through the link.

Regards
Bryan
User avatar
spacious_mind
Senior Member
Posts: 3999
Joined: Wed Aug 01, 2007 10:20 pm
Location: Alabama
Contact:

Post by spacious_mind »

Chessmaster Ireland wrote:Hi Nick

Very strange that.

I have just tested the video link on my phone and it takes me to the correct Amiga Sargon III video.

Maybe someone else can advise what video they can see through the link.

Regards
Bryan
OK it works the second time around.. that was weird.

He must be playing it Amiga 500 mode judging by the search depths :)

Anyway here is Sargon 3 on Mac:

Image

It came out in 1984. Fastest that I can get it go is about 8 ply which is around the same speed as Atari ST 68060-32 MHz emulator so its worth a test game between the two. Atari ST is playing in Division 3 already.

Here is Sargon 4 on Mac:

Image

Sargon 4 on Mac came out in 1988 same year as Amiga and Atari Sargon 3 so it is possible that Sargon 4 Mac is more similar to Amiga and Atari. It is supposed to not have all those bugs that the Sargon 4 DOS version has. Fastest I can get this to run is about 1 ply less than Amiga 68060 Sargon 3 so this is a worthy test too.

Mac Sargon 3 played together in the 1984 World Championship and finished behind Apple II but ahead of Compaq pc so at that time the 6502 Sargon 3 was better than the 68000 version. But then again the 68000 is slow.

Regards
Nick
User avatar
spacious_mind
Senior Member
Posts: 3999
Joined: Wed Aug 01, 2007 10:20 pm
Location: Alabama
Contact:

Post by spacious_mind »

Here is a game I played between Macintosh Classic Sargon 3 against Atari ST 68030-32 MHz Sargon 3.

[Event "Computer Test Match"]
[Site "Alabama"]
[Date "2017.12.09"]
[Round "1"]
[White "MAC68000 Sargon 3, 30S."]
[Black "AT68060-32 Sargon 3, 30S."]
[Result "1/2-1/2"]
[ECO "A03"]
[WhiteElo "2122"]
[BlackElo "2122"]
[Annotator ""]
[PlyCount "92"]
[EventDate "2017.12.09"]
[EventCountry "USA"]

1. f4 d5 2. Nf3 Nf6 3. e3 g6 4. Be2 Bg7 5. O-O O-O 6. d3 c5 7. Qe1 Nc6 8. Qh4 b6 9. Nbd2 Ba6 10. Ne5 Nxe5 11. fxe5 Nd7 {MAC68000 Sargon 3 out of book} 12. Qa4 {AT68060-32 Sargon 3 out of book} Qc8 13. Nf3 Nxe5 14. Bd2 Nxf3+ 15. Bxf3 Bb7 16. Rab1 Qc7 17. Qh4 f5 18. c4 d4 19. exd4 Bxf3 20. Rxf3 Bf6 21. Bg5 Bxd4+ 22. Kh1 e5 23. b3 Rfe8 24. Rff1 a5 25. Rbe1 a4 26. Bf6 axb3 27. axb3 Ra2 28. Qg3

[fen]4r1k1/2q4p/1p3Bp1/2p1pp2/2Pb4/1P1P2Q1/r5PP/4RR1K w - - 0 28[/fen]

Atari overall had the better game but missed a couple of opportunities to win this game. In the above position 28. ... Qc6 or f4 were good for Black.

Qd7 29. Bxe5 Bxe5 30. Rxe5 Rxe5 31. Qxe5 Qd8 32. Rf3 Qe7 33. Qb8+ Kg7 34. Rf1 Qe2 35. Rg1 Qxd3 36. Qxb6 Qd4 37. Qb7+ Kh6 38. Qf3 Ra1 39. Rf1 Kg7 40. Qe2 Qf6 41. Rd1 f4 42. Qd3 Qe5 43. Qd7+ Kh6 44. h3 Qf5 45. Qd8 Rxd1+ 46. Qxd1 Qe6 1/2-1/2

Final Position

[fen]8/7p/4q1pk/2p5/2P2p2/1P5P/6P1/3Q3K w - - 0 47[/fen]

Lichess Evaluation

https://lichess.org/8xnl2DTN

MAC68000 Sargon 3, 30S. (2122)
5 Inaccuracies
0 Mistakes
0 Blunders
18 Average centipawn loss

AT68060-32 Sargon 3, 30S. (2122)
4 Inaccuracies
1 Mistakes
0 Blunders
18 Average centipawn loss

Atari at 68060-32 MHz searched approximately as deep as MAC Classic Sargon 3 running on Max. So I guess you could say that both these Sargon 3 versions are pretty much the same. But it is hard to tell since Mac Sargon 3 came out in 1984 and Atari in 1988.

So next I tried Mac Classic Sargon 4 against Amiga 68060. Mac Sargon 4 played Max speed on my I7 6 core laptop and Amiga 68060 played on my 4.2 GHZ 8 core AMD desktop. The I7 is a little faster than the AMD 4.2 GHz.

Doing it this way, both programs searched almost the same depths.

[Event "Computer Test Match"]
[Site "Alabama"]
[Date "2017.12.09"]
[Round "1"]
[White "MAC68000 Sargon 4, 30S."]
[Black "AM68060 Sargon 3, 30S."]
[Result "0-1"]
[ECO "A05"]
[WhiteElo "2122"]
[BlackElo "2400"]
[Annotator ""]
[PlyCount "156"]
[EventDate "2017.12.09"]
[EventCountry "USA"]

1. Nf3 Nf6 2. g3 g6 3. b4 Bg7 4. Bb2 O-O 5. Bg2 c6 6. O-O a5 7. a3 axb4 8. axb4 Rxa1 9. Bxa1 Qb6 10. c3 d5 11. d3 Bg4 {MAC68000 Sargon 3 out of book} 12. Nbd2 {AM68060 Sargon 3 out of book} Nbd7 13. Qa4 e5 14. e4 h6 15. Bb2 Be6 16. Ra1 Ng4 17. Qa7 Qc7 18. exd5 cxd5 19. Ra5 Qc8 20. c4 b6 21. Ra1 dxc4 22. dxc4 Rd8 23. h3 Ngf6 24. Rc1 Nh7 25. Qa3 Qc7 26. Qc3 f5 27. Re1 Bf7

[fen]3r2k1/2qn1bbn/1p4pp/4pp2/1PP5/2Q2NPP/1B1N1PB1/4R1K1 w - - 0 28[/fen]

[fen]3r2k1/2qn1bbn/1p4pp/4pp2/1PP5/2Q2NPP/1B1N1PB1/4R1K1 w - - 0 28[/fen]

After 27 moves the position is equal. However 28. Kh2 was played in the above position which quickly gives Black advantage that Amiga Sargon 3 also took advantage of playing the move 28. ... Nhf6.

28. Kh2 Nhf6 29. Qc2 Nh5 30. Kg1 e4 31. Nh2 Bxb2 32. Qxb2 Ne5 33. Qc3 Rd3 34. Qc2 Nxc4 35. Nxc4 Qxc4 36. Qxc4 Bxc4 37. Bf1 Rd4 38. Bxc4+ Rxc4 39. Rb1 Nf6 40. b5 Nd5 41. f3 e3 42. f4 Rc2 43. Re1 Rb2 44. h4 h5 45. Nf1 e2 46. Nh2 Nb4 47. Nf3 Nd3 48. Ra1 Kf7 49. Ne1 Nc5 50. Ra7+ Ke6 51. Kf2 Ne4+ 52. Kf3 Nc3 53. Rg7 Kf6 54. Rc7 Nxb5 55. Rc6+ Ke7 56. Rxg6 Nc3 57. Rg7+ Kf6 58. Ra7 b5 59. Ra6+ Kg7 60. Ra7+ Kg6 61. Ra6+ Kf7 62. Ra7+ Ke6 63. Ra6+ Kd5 64. Ra8 b4 65. Rd8+ Ke6 66. Rc8 Ke7 67. Rc6 Kd7 68. Rc4 Ne4 69. Rd4+ Ke6 70. Rd8 Ra2 71. Re8+ Kd7 72. Re5 b3 73. Rb5 b2 74. Kg2 Ra1 75. Rxb2 Rxe1 76. Kf3 Rf1+ 77. Kxe2 Rf2+ 78. Kd3 Rxb2 0-1

Final Position

[fen]8/3k4/8/5p1p/4nP1P/3K2P1/1r6/8 w - - 0 79[/fen]

Mac Classic Sargon 4 worked hard but it was a losing battle. Amiga 68060 Sargon 3 wins.

Lichess Evaluation

https://lichess.org/yXcM3lhJ

MAC68000 Sargon 4, 30S. (2122)
9 Inaccuracies
5 Mistakes
1 Blunders
30 Average centipawn loss

AM68060 Sargon 3, 30S. (2400)
3 Inaccuracies
1 Mistakes
0 Blunders
17 Average centipawn loss

Well after these tests it seems that Sargon 3 on a WinUAE Amiga Emulator running at full speed on an I7 is still the fastest and possibly best Sargon version.

Mac Classic Sargon 4 plays approximately 9 ply deep on an I7 with its 68,000 program. There is an emulator called Basilisc II for Mac but its very fiddly to make work but you should be able to get a Power Mac running on it. If someone has ever had it working then please let me know how.

I do however considering how fiddly it is think that WINUae Amiga probably remain the deepest searcher.

Classic Mac Sargon 4 searches deeper than Classic Mac Sargon 3, therefore for sure the search speed was improved on Sargon 4.

But, since Amiga Sargon 3 with matching search depth still seems to be probably slightly better than Sargon 4. You can make a lot of assumptions.

You could assume.

1) Mac Sargon 3 from 1984 = early version = maybe the same as Atari ST and Amiga from 1988 but maybe less likely.
2) Mac Sargon 4 from 1988 = maybe the same as Atari ST and Amiga from 1988
3) Maybe Amiga Sargon 3 is the best Sargon 3 conversion when playing at same search speed as the faster searcher Sargon 4, slower searcher Sargon 3 is better. Same could apply to Atari Sargon 3.
4) Maybe Amiga Sargon 3 is the best conversion plain and simple.

But having said all that I think as a result of these games that MAC Classic Sargon 3 is good enough to compete in Division 3. Mac Classic Sargon 4 just like Amiga 68060 Sargon 3 I would expect to be competing near the top of the Division 3 table. I will probably include it in Division 2 in the future.

Also what is certain is that the Mac Sargon 3 and Sargon 4 conversions just like Amiga and Atari yet again embarrass the poor PC conversions that were sold.

Best regards
Nick
User avatar
Bryan Whitby
Senior Member
Posts: 1001
Joined: Wed Feb 18, 2009 9:57 pm
Location: England

Post by Bryan Whitby »

Great post Nick

Very interesting and informative, thanks for all your time and effort.

With the snow falling all day yesterday, I was stuck indoors for most of the day. After reading in the morning about Sargon, I fired up the old PC in the afternoon and battled against a variety of Fidelity MESS chess computers. The great news was that I actually won one game!! The poor sucker that didn't know how to play a completely drawn endgame was the mighty Playmatic S.

Too embarrassed to post winning game.

Regards
Bryan
User avatar
Bryan Whitby
Senior Member
Posts: 1001
Joined: Wed Feb 18, 2009 9:57 pm
Location: England

Post by Bryan Whitby »

Nick

Forgot to say that I played on my special MESS Chess Board that was kindly programmed by my good friend Berger.

Photo is me playing one of the Novag's (Scotch Opening)

Bryan

Image
Post Reply