Nodes/sec. with last new CPU's!

[Vinvin]

Regarding spassky vs fischer position above, H6 Pro on 8-Core i7-5960X 4.1GHz (5-men syzygy) evaluate position after 30. g3 as +0.25

Maybe if someone analyse with Xeon, H6 will finally evaluate the position as triple zero!

That would be groundbreaking, the only deep analysis I saw always show wins for white.

[Nordlandia]

Regarding spassky vs fischer position above, H6 Pro on 8-Core i7-5960X 4.1GHz (5-men syzygy) evaluate position after 30. g3 as +0.25

Maybe if someone analyse with Xeon, H6 will finally evaluate the position as triple zero!

That would be groundbreaking, the only deep analysis I saw always show wins for white.

High score for white is because engine does not recognize fortress as draw.

One line that leads to draw.

[pgn][Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "New game"]
[Black "?"]
[Result "*"]
[SetUp "1"]
[FEN "5k2/pp4pp/4pp2/1P6/8/P2KP1P1/5P1b/2B5 b - - 0 0"]
[PlyCount "43"]

1... h5 2. Ke2 h4 3. Kf3 h3 4. Kg4 Bg1 5. Kxh3 Bxf2 6. Bd2 a6 7. b6 Ke7 8. Kg2
Bxg3 9. Kxg3 g6 10. Bb4+ Kd7 11. Kf4 e5+ 12. Ke4 Ke6 13. a4 f5+ 14. Kf3 Kd5 15.
a5 g5 16. e4+ fxe4+ 17. Ke3 Kc6 18. Kxe4 Kd7 19. Kxe5 Kc8 20. Bd6 Kd7 21. Bc7
Kc8 22. Kf5 Kd7 *

[/pgn]

I'm able to drew H6 as black!

[pgn][Event "Blitz 2m"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "Houdini 6 Pro x64-pext"]
[Black "Kajapix"]
[Result "1/2-1/2"]
[SetUp "1"]
[FEN "5k2/pp4pp/4pp2/1P6/8/P2KP1P1/5P1b/2B5 b - - 0 1"]
[PlyCount "126"]
[TimeControl "120"]

{1024MB, Fritz 15.ctg, EXPERT-PC} 1... h5 {0} 2. Ke2 {0.41/26 3} h4 {2} 3. Kf3
{0.62/27 0} h3 {1 (hxg3)} 4. Kg4 {0.58/30 3} Bg1 {0} 5. Kxh3 {0.69/29 2} Bxf2 {
0} 6. Bd2 {0.51/28 4} a6 {0 (Ke7)} 7. b6 {0.77/28 1} Ke7 {2} 8. Kg2 {0.83/30 1}
Bxg3 {2 (Bxe3)} 9. Kxg3 {1.26/24 2} g6 {0 (Kf7)} 10. a4 {1.65/28 2} e5 {1} 11.
a5 {2.24/28 6} Ke6 {3 (Kd6)} 12. Kg4 {2.37/30 3} Kf7 {3 (Kd7)} 13. Kf3 {
2.49/28 2} Ke6 {1} 14. Ke4 {2.44/27 0} Kd6 {2} 15. Bb4+ {2.57/27 0} Ke6 {1} 16.
Bc5 {2.68/29 1} g5 {7 (f5+)} 17. Kf3 {2.73/37 3} f5 {1} 18. e4 {2.73/38 0}
fxe4+ {1 (g4+)} 19. Kxe4 {2.74/39 3} Kd7 {0} 20. Bf2 {2.74/38 1} Kc8 {0 (Ke6)}
21. Kxe5 {2.74/44 3} Kb8 {0 (Kd7)} 22. Bg3 {2.74/47 17} Ka8 {0 (g4)} 23. Kf5 {
2.74/38 14} g4 {2} 24. Kf4 {2.74/34 0} Kb8 {6} 25. Kxg4+ {2.74/35 5} Ka8 {2}
26. Kf4 {2.74/34 8} Kb8 {0} 27. Kf3+ {2.74/32 1} Ka8 {0 (Kc8)} 28. Bf2 {
2.74/36 8} Kb8 {0} 29. Be3 {2.74/33 6} Ka8 {0 (Kc8)} 30. Kf4 {2.74/31 5} Kb8 {0
} 31. Bd4 {2.74/33 3} Ka8 {3} 32. Ke4 {2.74/30 0} Kb8 {1} 33. Bf6 {2.74/28 2}
Ka8 {0 (Kc8)} 34. Kf3 {2.74/25 2} Kb8 {0} 35. Bc3 {2.74/25 2} Ka8 {0 (Kc8)} 36.
Kf4 {2.74/22 0} Kb8 {0} 37. Bf6 {2.74/23 0} Ka8 {0 (Kc8)} 38. Ke4 {2.74/20 0}
Kb8 {0} 39. Be5+ {2.74/19 1} Ka8 {0 (Kc8)} 40. Bb2 {2.74/16 0} Kb8 {0} 41. Bd4
{2.74/16 0} Ka8 {0} 42. Ke3 {2.74/15 0} Kb8 {0} 43. Kf2 {2.74/14 0} Ka8 {
0 (Kc8)} 44. Kf1 {2.66/12 1} Kb8 {0} 45. Ke2 {2.74/12 0} Ka8 {0 (Kc8)} 46. Kd3
{2.74/10 0} Kb8 {0} 47. Ke3 {2.74/22 0} Ka8 {0} 48. Bc5 {0.82/23 0} Kb8 {0} 49.
Ke4 {0.90/20 0} Ka8 {0 (Kc8)} 50. Kd5 {0.80/16 0} Kb8 {0} 51. Kd6 {1.23/21 0}
Ka8 {0 (Kc8)} 52. Kd7 {0.99/16 0} Kb8 {0} 53. Ke7 {0.82/27 0} Ka8 {0 (Kc8)} 54.
Bb4 {0.44/18 0} Kb8 {0} 55. Kd7 {0.70/28 0} Ka8 {0} 56. Be1 {0.00/29 0} Kb8 {0}
57. Bd2 {0.00/44 0} Ka8 {0} 58. Bc1 {0.00/59 0} Kb8 {0} 59. Bb2 {0.00/70 0} Ka8
{0} 60. Ba1 {0.00/95 0} Kb8 {0} 61. Bb2 {0.00/99 0} Ka8 {0} 62. Ba1 {0.00/99 0}
Kb8 {0} 63. Bb2 {0.00/99 0} Ka8 {0} 64. Ba1 {0.00/99 0} 1/2-1/2

[/pgn]

[jpqy]

A bigger update ,also with charts

http://www.ipmanchess.yolasite.com/amd- ... -bench.php

JP.

[Leo]

So Fischer resigned in a drawn position.

[Geonerd]

A bigger update ,also with charts

http://www.ipmanchess.yolasite.com/amd- ... -bench.php

JP.

Excellent!

[cdani]

Awesome!!

[tpoppins]

Here under they used asmFishL 2017-09-19 Base version ,so you will see lower nodes/sec. but still a nice compair!

The latest asmFishL commit (rev 742) is dated 2017-08-25.
So what's a "compair", then? Hoax, typo or PEBCAK?

[Geonerd]

Greetings, Vael.

Here's another system for you.

29,335,688 NPS

2x Opteron 6176 = 24 cores, 24 threads, at 2.3 GHz.
05/22/17 ASMFish, POPCNT.
Supermicro H8DGU, 32GB DDR3-1333/10600. All 8 memory channels populated.
Win 7

Faster than an 8/16 Ryzen, for less than $200 U.S.!

No bios or driver optimization yet.

Hoping for more with OC Bios and perhaps memory system tweaks.

Thanks for maintaining your list!