Four examples comparing HT

[kgburcham]

I know there has been a lot said about chess programs using HT.
After reading some recent posts, I thought I would test this again.
Ran all four tests to depth 30 from the start position.

all tests split depth 12
first two tests are with HT on
second two tests are with HT off
SR-2 system at 4gig

[D] rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq -

CPU0: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU1: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU2: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU3: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU4: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU5: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU6: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU7: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU8: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU9: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU10: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU11: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU12: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU13: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU14: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU15: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU16: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU17: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU18: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU19: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU20: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU21: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU22: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU23: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
24 processor(s) found, POPCNT available
NUMA configuration with 2 node(s), offset 0
4096 MB Large Page Hash
Nalimov 6 men EGTB available - 999 MB cache
Engine: Houdini 3 Pro x64 (4096 MB)
by Robert Houdart
17/42 0:00 +0.16 1.d4 d5 2.Nc3 Nf6 3.Bf4 Bf5 4.Nf3 e6 5.e3 Nc6 6.Bd3 Bd6 7.O-O O-O 8.Bxf5 exf5 9.Bxd6 Qxd6 10.Nb5 Qe7 11.Qd3 (18.404.127) 20336
18/42 0:01 +0.16 1.d4 Nf6 2.Nc3 d5 3.Bf4 Bf5 4.Nf3 e6 5.e3 Nc6 6.Bd3 Bd6 7.Bxd6 Qxd6 8.O-O O-O 9.Bxf5 exf5 10.Nb5 Qe7 11.Qd3 (27.547.166) 21915
18/42 0:01 +0.23 1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.e5 Qe7 7.Qe2 Nd5 8.c4 Ba6 9.a3 Nb6 10.b3 d5 11.exd6 cxd6 (30.845.080) 22111
19/42 0:01 +0.21 1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.e5 Qe7 7.Qe2 Nd5 8.c4 Qb4+ 9.Nd2 Nf4 10.Qe4 Ng6 11.a3 Qa5 (35.448.218) 22421
20/42 0:01 +0.21 1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.e5 Qe7 7.Qe2 Nd5 8.c4 Qb4+ 9.Nd2 Nf4 10.Qe4 Ng6 11.a3 Qa5 (44.122.612) 23234
21/42 0:02 +0.21 1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.e5 Qe7 7.Qe2 Nd5 8.c4 Qb4+ 9.Nd2 Nf4 10.Qe4 Ng6 11.a3 Qa5 (74.802.412) 25025
22/44 0:07 +0.19 1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.Bg5 dxe4 5.Nxe4 Be7 6.Nxf6+ Bxf6 7.Bxf6 Qxf6 8.Nf3 O-O 9.c3 Bd7 10.Ne5 Bc6 11.Qg4 g6 (227.253.221) 30099
23/44 0:12 +0.23 1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.Bg5 dxe4 5.Nxe4 Be7 6.Nxf6+ Bxf6 7.Bxf6 Qxf6 8.Nf3 O-O 9.c3 Bd7 10.Ne5 Bc6 11.Qg4 g6 (391.144.351) 31216
24/49 0:18 +0.23 1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.Bg5 dxe4 5.Nxe4 Nbd7 6.Nf3 h6 7.Nxf6+ Nxf6 8.Bxf6 Qxf6 9.Bd3 Bd6 10.O-O O-O 11.Qe2 Qf4 (596.093.075) 31847
25/49 0:27 +0.23 1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.Bg5 dxe4 5.Nxe4 Nbd7 6.Nf3 h6 7.Nxf6+ Nxf6 8.Bxf6 Qxf6 9.Bd3 Bd6 10.O-O O-O 11.Qe2 Qf4 (901.242.606) 32542
26/52 0:43 +0.15-- 1.e4 e6 (1.451.366.560) 33577
26/58 2:14 +0.17 1.d4 e6 2.Nf3 Nf6 3.c4 Bb4+ 4.Bd2 Qe7 5.Nc3 O-O 6.e3 d6 7.Qc2 b6 8.Bd3 Bb7 9.O-O Bxf3 10.gxf3 c5 11.a3 Bxc3 (4.723.478.024) 35008
27/58 2:48 +0.21 1.d4 e6 2.Nf3 Nf6 3.c4 Bb4+ 4.Bd2 Qe7 5.Nc3 O-O 6.e3 d6 7.Qc2 b6 8.Bd3 Bb7 9.O-O Bxf3 10.gxf3 c5 11.dxc5 bxc5 (5.877.376.623) 34948
28/58 3:52 +0.19 1.d4 e6 2.Nf3 Nf6 3.c4 Bb4+ 4.Bd2 Qe7 5.a3 Bxd2+ 6.Qxd2 b6 7.Nc3 Bb7 8.Qg5 O-O 9.e4 h6 10.Qe3 d6 11.e5 Nfd7 (8.153.139.001) 35114
29/66 9:39 +0.16 1.d4 Nf6 2.c4 e6 3.Nf3 Bb4+ 4.Bd2 Qe7 5.a3 Bxd2+ 6.Qxd2 b6 7.Nc3 Bb7 8.Qg5 O-O 9.e4 h6 10.Qe3 d6 11.e5 Nfd7 (20.863.146.577) 35996
30/66 16:40 +0.22 1.d4 Nf6 2.c4 e6 3.Nf3 Bb4+ 4.Bd2 Qe7 5.g3 Bxd2+ 6.Qxd2 Nc6 7.Nc3 d5 8.Bg2 dxc4 9.Rd1 Qb4 10.O-O O-O 11.e3 Rb8 (36.119.515.395) 36083

repeat of first test
2nd test
HT on
Engine: Houdini 3 Pro x64 (4096 MB)
by Robert Houdart
19/34 0:00 +0.23 1.d4 Nf6 2.Nc3 d5 3.Bf4 e6 4.e3 Bd6 5.Nf3 Nc6 6.Bd3 O-O 7.O-O Bd7 8.a3 Bxf4 9.exf4 h6 10.Qe2 Qe7 11.Nb5 (15.483.805) 21595
20/37 0:01 +0.21 1.d4 Nf6 2.Nc3 d5 3.Bf4 e6 4.e3 Bd6 5.Nf3 Nc6 6.Bd3 O-O 7.O-O Bd7 8.a3 Bxf4 9.exf4 h6 10.Qd2 Qe7 11.h3 (26.406.683) 23725
21/40 0:01 +0.18 1.d4 Nf6 2.Nc3 d5 3.Bf4 e6 4.e3 Bd6 5.Nf3 Nc6 6.Bd3 O-O 7.O-O Bd7 8.a3 h6 9.h3 a6 10.Qd2 Qe7 11.Bxd6 cxd6 (50.015.640) 26199
21/42 0:03 +0.26 1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.O-O Be7 5.Nc3 O-O 6.Bxc6 dxc6 7.Nxe5 Nxe4 8.Nxe4 Qd4 9.Nxc6 bxc6 10.Re1 Rb8 11.d3 Re8 (85.505.868) 28473
22/49 0:08 +0.18-- 1.e4 c5 (260.110.698) 31935
22/49 0:09 +0.19 1.e4 c5 2.Nc3 Nc6 3.Bb5 e5 4.Nf3 d6 5.a3 Nf6 6.b4 Qb6 7.d3 a6 8.Bc4 Bg4 9.O-O Nd4 10.Rb1 Qc7 11.Nd5 Nxd5 (299.881.595) 31477
23/49 0:13 +0.27++ 1.e4 (434.011.712) 31347
23/49 0:16 +0.23 1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 e5 6.Bb5+ Bd7 7.Bxd7+ Qxd7 8.Nde2 Nc6 9.O-O Be7 10.Bg5 O-O (524.948.126) 31507
24/49 0:19 +0.23 1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 e5 6.Bb5+ Bd7 7.Bxd7+ Qxd7 8.Nde2 Nc6 9.O-O Be7 10.Bg5 O-O 11.Bxf6 Bxf6 (624.034.179) 31612
25/51 0:30 +0.22 1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Qxd4 e5 5.Qd3 Nf6 6.Nc3 Be7 7.Be2 O-O 8.O-O Nc6 9.a3 Be6 10.Be3 h6 11.Rad1 Qd7 (977.789.725) 32446
26/56 0:55 +0.22 1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Qxd4 e5 5.Qd3 Nf6 6.Nc3 Be7 7.Be2 O-O 8.O-O Nc6 9.a3 Be6 10.Be3 h6 11.Rad1 Qd7 (1.857.736.803) 33678
27/57 1:31 +0.19 1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Qxd4 Nc6 5.Qe3 Nf6 6.Nc3 g6 7.h3 Bg7 8.Be2 O-O 9.O-O Nb4 10.Qd2 Nc6 11.a3 Nd7 (3.150.982.399) 34358
28/57 2:31 +0.20 1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 Nc6 6.Bb5 Bd7 7.O-O g6 8.Nxc6 bxc6 9.Bc4 Bg7 10.h3 O-O 11.Be3 Rb8 (5.275.723.358) 34851
29/69 7:39 +0.19 1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3 e6 7.Be2 Be7 8.Qd2 O-O 9.O-O-O Qc7 10.Kb1 b5 11.f3 b4 (16.520.656.057) 35986
30/69 16:26 +0.11-- 1.e4 e6 (36.157.301.196) 36661

3rd test HT off
CPU0: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU1: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU2: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU3: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU4: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU5: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU6: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU7: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU8: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU9: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU10: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
CPU11: GenuineIntel Intel64 Family 6 Model 44 Stepping 0 4003 MHz
12 processor(s) found, POPCNT available
NUMA configuration with 2 node(s), offset 0
4096 MB Large Page Hash
Nalimov 6 men EGTB available - 999 MB cache
Engine: Houdini 3 Pro x64 (4096 MB)
by Robert Houdart
18/35 0:00 +0.21 1.d4 Nf6 2.Nf3 d5 3.Bf4 Bf5 4.e3 e6 5.Bd3 Bd6 6.Bxf5 exf5 7.O-O O-O 8.Qd3 Bxf4 9.exf4 Ne4 10.Qb5 Nc6 11.Qxb7 (20.625.578) 21620
18/35 0:01 +0.24++ 1.e4 (25.773.552) 21991
18/37 0:01 +0.23 1.e4 e5 2.Nf3 Nc6 3.Bb5 Bc5 4.O-O Nf6 5.c3 a6 6.Bxc6 dxc6 7.d4 exd4 8.cxd4 Be7 9.Nc3 O-O 10.Qd3 Be6 11.Bf4 h6 (28.197.892) 22029
19/40 0:01 +0.31++ 1.e4 (35.255.357) 22527
19/40 0:01 +0.30 1.e4 e5 2.Nf3 Nc6 3.Bb5 Bc5 4.O-O d6 5.c3 a6 6.Bxc6+ bxc6 7.d4 exd4 8.cxd4 Ba7 9.h3 Ne7 10.Nc3 O-O 11.Qd3 Rb8 (37.333.612) 22382
20/40 0:01 +0.22-- 1.e4 e5 (43.056.514) 22578
20/40 0:02 +0.12 1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.O-O Nxe4 5.Re1 Nd6 6.Nxe5 Be7 7.Bf1 Nxe5 8.Rxe5 O-O 9.d4 Bf6 10.Qf3 h6 11.Nc3 c6 (54.639.758) 23045
20/40 0:02 +0.16 1.d4 Nf6 2.Nc3 d5 3.Bf4 a6 4.e3 Nc6 5.Nf3 e6 6.Bd3 Bd6 7.O-O O-O 8.Qd2 h6 9.h3 Bd7 10.e4 Nb4 11.e5 Nxd3 12.Qxd3 (66.703.920) 23553
21/40 0:04 +0.16 1.d4 Nf6 2.Nc3 d5 3.Bf4 Bf5 4.e3 e6 5.Bd3 Bxd3 6.cxd3 Bd6 7.Nge2 Nc6 8.O-O O-O 9.Qb3 Rb8 10.a3 Qd7 11.Rac1 Bxf4 12.Nxf4 (107.985.015) 24767
22/40 0:05 +0.17 1.d4 Nf6 2.Nc3 d5 3.Bf4 e6 4.e3 Nc6 5.Bd3 a6 6.Nf3 Bd6 7.O-O O-O 8.Qd2 Bxf4 9.exf4 Bd7 10.Ne5 Qe7 11.Ne2 Nxe5 (141.647.133) 25132
23/45 0:08 +0.17 1.d4 Nf6 2.Nc3 d5 3.Bf4 e6 4.e3 Nc6 5.Bd3 a6 6.Nf3 Bd6 7.O-O O-O 8.Qd2 h6 9.h3 Bxf4 10.exf4 Qe7 11.a3 Bd7 12.Rae1 Rfe8 (219.840.313) 25839
24/52 0:23 +0.14 1.d4 Nf6 2.Nf3 d5 3.c4 e6 4.Nc3 Be7 5.cxd5 exd5 6.Bf4 Nc6 7.e3 O-O 8.Bd3 Nh5 9.O-O Nxf4 10.exf4 g6 11.Re1 Be6 (627.874.851) 26638
25/54 0:43 +0.22++ 1.d4 (1.208.781.225) 27528
25/54 1:16 +0.18 1.d4 d5 2.c4 e6 3.Nf3 a6 4.c5 Nc6 5.Bf4 Nf6 6.e3 Nh5 7.Bg5 Be7 8.Bxe7 Qxe7 9.Nc3 Nf6 10.Be2 O-O 11.O-O Bd7 12.Qb3 Rab8 (2.158.445.716) 28274
26/54 2:00 +0.23 1.d4 d5 2.c4 e6 3.Nf3 Nf6 4.Nc3 h6 5.g3 Nc6 6.Bg2 Bd6 7.O-O O-O 8.c5 Be7 9.Rb1 a6 10.a3 b5 11.b4 Bb7 12.Qd3 Qd7 (3.429.863.936) 28576
27/64 3:06 +0.22 1.d4 Nf6 2.c4 c5 3.d5 b5 4.cxb5 a6 5.b6 Qxb6 6.Nc3 d6 7.e4 Nbd7 8.Nf3 g6 9.h3 Bg7 10.Bd3 Nh5 11.O-O O-O (5.393.481.055) 28876
28/64 5:04 +0.17 1.d4 Nf6 2.c4 c5 3.d5 b5 4.cxb5 a6 5.b6 Qxb6 6.Nc3 d6 7.e4 Nbd7 8.Nf3 g6 9.h3 Bg7 10.Bd3 Nh5 11.O-O O-O 12.Rb1 Ne5 (8.893.858.892) 29224
29/64 8:49 +0.20 1.d4 Nf6 2.c4 c5 3.d5 b5 4.cxb5 a6 5.b6 Qxb6 6.Nc3 d6 7.e4 Nbd7 8.Nf3 g6 9.h3 Bg7 10.Rb1 O-O 11.Be2 Bb7 12.O-O a5 (15.716.946.933) 29665
30/64 12:53 +0.19 1.d4 Nf6 2.c4 c5 3.d5 b5 4.cxb5 a6 5.b6 Qxb6 6.Nc3 d6 7.e4 Nbd7 8.Nf3 g6 9.h3 Bg7 10.Rb1 O-O 11.Be2 a5 12.O-O Ba6 (23.153.540.638) 29915

repeat of third test
fourth test
HT off
Engine: Houdini 3 Pro x64 (4096 MB)
by Robert Houdart
16/37 0:00 +0.14 1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.e5 Qe7 7.Qe2 Nd5 8.c4 Ba6 9.a3 f6 10.exf6 Nxf6 11.Nc3 d5 12.Bf4 Bxc4 (6.157.979) 6911
16/37 0:00 +0.18 1.d4 d5 2.Nc3 Bf5 3.Bf4 e6 4.e3 Nc6 5.Nf3 Bd6 6.Bxd6 cxd6 7.Bd3 Nf6 8.Bxf5 exf5 9.O-O O-O (6.887.258) 7389
17/40 0:01 +0.19 1.d4 d5 2.Nc3 Nf6 3.Bf4 e6 4.e3 Bd6 5.Nf3 Nc6 6.Be2 Bxf4 7.exf4 Bd7 8.O-O O-O 9.Ne5 Nxe5 10.fxe5 Ne4 11.Nxe4 dxe4 (11.247.733) 9927
18/40 0:01 +0.11-- 1.d4 d5 (14.580.857) 11373
18/40 0:01 +0.14 1.d4 d5 2.Nc3 Nf6 3.Bf4 e6 4.e3 Bd6 5.Nf3 Nc6 6.Be2 Bxf4 7.exf4 Qd6 8.O-O Qxf4 9.g3 Qd6 10.Nb5 Qe7 11.Ne5 Nxe5 12.dxe5 (21.008.539) 13364
19/40 0:01 +0.22++ 1.d4 (23.647.758) 14143
19/40 0:01 +0.19 1.d4 d5 2.Nc3 Nf6 3.Bf4 e6 4.e3 Bd6 5.Nf3 Nc6 6.Be2 O-O 7.O-O Bd7 8.Ne5 h6 9.Qd3 Bxe5 10.Bxe5 Nxe5 11.dxe5 (24.556.059) 14368
20/41 0:02 +0.16 1.d4 d5 2.Bf4 Nc6 3.e3 Bf5 4.Nc3 e6 5.Bd3 Bxd3 6.cxd3 Bd6 7.Nge2 Nf6 8.O-O O-O 9.Qb3 Bxf4 10.Nxf4 Rb8 11.Rac1 Qe7 12.a3 (47.132.164) 18475
21/41 0:03 +0.16 1.d4 d5 2.Bf4 Nc6 3.e3 Bf5 4.Nc3 e6 5.Bd3 Bxd3 6.cxd3 Bd6 7.Nge2 Nf6 8.O-O O-O 9.Qb3 Bxf4 10.Nxf4 Rb8 11.Rac1 Qe7 12.a3 (67.139.193) 20053
22/41 0:05 +0.15 1.d4 d5 2.Bf4 Nc6 3.e3 Bf5 4.Nc3 e6 5.Bd3 Bxd3 6.cxd3 Bd6 7.Nge2 Nge7 8.O-O O-O 9.Rc1 Nf5 10.Qb3 Rb8 11.Nb5 Qe7 (126.995.667) 23023
22/41 0:08 +0.24++ 1.e4 (200.051.434) 24567
22/49 0:12 +0.28 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 5.O-O Qf6 6.d3 Bd6 7.c4 b6 8.Nc3 h6 9.Qa4 Bd7 10.Bd2 Qe6 11.d4 exd4 12.Nxd4 Qe5 (317.046.089) 25786
23/50 0:16 +0.20-- 1.e4 c5 (442.550.914) 26237
23/50 0:19 +0.19 1.e4 c5 2.Nf3 Nc6 3.Bb5 g6 4.Nc3 Bg7 5.h3 Nd4 6.Bc4 d6 7.d3 Nf6 8.O-O O-O 9.a4 a6 10.a5 Be6 11.Bxe6 Nxe6 12.Re1 Rc8 (501.796.244) 25548
24/50 0:23 +0.20 1.e4 c5 2.Nf3 Nc6 3.Bb5 g6 4.Nc3 Bg7 5.O-O Nf6 6.Re1 O-O 7.e5 Ne8 8.d3 Nc7 9.Bc4 Ne6 10.a3 Ncd4 11.Be3 d6 12.exd6 exd6 (599.323.213) 25860
25/52 0:42 +0.13 1.e4 e6 2.d4 d5 3.Nd2 Nf6 4.e5 Nfd7 5.Bd3 c5 6.c3 Nc6 7.Ne2 cxd4 8.cxd4 f6 9.exf6 Nxf6 10.O-O Bd6 11.Nf3 O-O (1.166.644.549) 27424
25/56 1:32 +0.24 1.d4 d5 2.c4 e6 3.Nf3 Nf6 4.Nc3 h6 5.Bf4 Bd6 6.Bxd6 Qxd6 7.e3 O-O 8.Be2 Nc6 9.O-O dxc4 10.Bxc4 a6 11.a3 Rd8 12.Qd3 b5 (2.639.170.006) 28442
26/56 1:51 +0.21 1.d4 d5 2.c4 e6 3.Nf3 Nf6 4.Nc3 h6 5.g3 dxc4 6.Qa4+ Bd7 7.Qxc4 Bc6 8.Bg2 Bd6 9.O-O O-O 10.Re1 Ne4 11.Qd3 f5 12.a3 Nxc3 (3.162.029.103) 28464
27/56 2:15 +0.21 1.d4 d5 2.c4 e6 3.Nf3 Nf6 4.Nc3 h6 5.g3 dxc4 6.Qa4+ Bd7 7.Qxc4 Bc6 8.Bg2 Bd6 9.O-O O-O 10.Re1 Ne4 11.Qd3 f5 12.Nd2 Nxd2 (3.875.042.120) 28615
28/56 3:18 +0.21 1.d4 d5 2.c4 e6 3.Nf3 Nf6 4.Nc3 h6 5.g3 Bd6 6.Bg2 O-O 7.O-O Nc6 8.c5 Be7 9.Rb1 a6 10.Bf4 b6 11.b4 Bb7 12.a3 Nh5 (5.762.242.963) 29078
29/63 8:09 +0.20 1.d4 Nf6 2.c4 e6 3.Nf3 Bb4+ 4.Bd2 Qe7 5.e3 O-O 6.Be2 b6 7.O-O Bb7 8.Bxb4 Qxb4 9.Qc2 Be4 10.a3 Qe7 11.Qd2 Nc6 12.Nc3 Bg6 (14.637.948.372) 29920
30/63 12:24 +0.19 1.d4 Nf6 2.c4 e6 3.Nf3 Bb4+ 4.Bd2 Qe7 5.e3 O-O 6.Be2 b6 7.Bxb4 Qxb4+ 8.Qd2 Qxd2+ 9.Nfxd2 Bb7 10.O-O Nc6 11.Nc3 h6 12.Bf3 a6 (22.456.785.197) 30150

[yanquis1972]

30Mn/s!! nice machine.

imo, not enough tests to conclude anything. i cherry-picked depth 27 to examine, & the hyperthreaded version faster. i then looked at d29, and its about even, with the fastest time to depth being w/ HT on. if i split the difference & look at depth 28, similar results: 3min & 5min w/ HToff, 4 min & 2.5 min w/ HT on.

seems possible to me the hyperthreaded runs possibly happened to have 'bad luck' between depths 29 & 30, but of course, the longer & deeper the search, the more accurate the results will be & it's equal possible that the ineffeciences take thaat long to become apparent.

[Laskos]

See my test on 100 positions, repeated 3 times
http://talkchess.com/forum/viewtopic.ph ... &start=140
Hypethreading gives in my case 12% speedup on time to depth, but if you are overclocking, it heats more.

[mwyoung]

See my test on 100 positions, repeated 3 times
http://talkchess.com/forum/viewtopic.ph ... &start=140
Hypethreading gives in my case 12% speedup on time to depth, but if you are overclocking, it heats more.

I got about the same result on my system. Around 10%. I don't argue this point with anyone anymore. They can run their system and test how every they wish. My test have always shown a speed up, and at worst no effect at all depending on the program. Until someone can prove that my eyes are lying to me when doing this test. I am running with HT on and using all the cores. The effect is not huge, but is measurable.

[yanquis1972]

i too have this impression. right now i don't have adequate cooling & thus only use physical cores, but i 'felt' (as i never ran any kind of true study) that HT was in fact beneficial, regardless of the increased nodecount.

i've never understood why engine authors are almost universally opposed to the idea. i would assume it's because they have a deeper understanding, but it might be that the 'noise'/inefficiency they see with hyperthreading (invisible to me) influences their opinion more.

[mwyoung]

i too have this impression. right now i don't have adequate cooling & thus only use physical cores, but i 'felt' (as i never ran any kind of true study) that HT was in fact beneficial, regardless of the increased nodecount.

i've never understood why engine authors are almost universally opposed to the idea. i would assume it's because they have a deeper understanding, but it might be that the 'noise'/inefficiency they see with hyperthreading (invisible to me) influences their opinion more.

I have done to many test to count trying to show the authors are correct. I see no harm in using HT, remember some say it hurts the engine. That would be easier to see, not harder.IMO

I ignore NPS it means nothing, time to depth only. I have used low depth and very high depth in testing this. All show it is better to use HT and max cores on my system. So that is what I use.

[Jhoravi]

Maybe Time to depth is not a good comparison because every core added must have changed the search flow resulting to same depth but different strength. I propose "Time to Solve"

[carldaman]

Maybe Time to depth is not a good comparison because every core added must have changed the search flow resulting to same depth but different strength. I propose "Time to Solve"

As pointed out in this forum before, difficult tactical solutions can be obtained faster with HT on, because it results in better move ordering. However, in more normal, quiet, less tactical positions, HT has a detrimental effect. At least, this has been the accepted theory.

We still need better proof to back up these claims that using HT results in greater overall strength. Solving difficult tactical problems is not the best way to test for this, since such positions do not represent the majority of chess positions.

CL

[IWB]

..., but if you are overclocking, it heats more.

If you are overclocking you get more than 10% extra clock when HT is off and now you have the same problem again.

Besides that, HT when using all core with one engine might(!) be ok, when using more than one engine at a time (worst is just single core engines) you run into problems as the load is not nessesarily distributed equaly.

The best would certainly be if there would be no HT in CPU. Things would be clear and easy!

Bye
Ingo

[Jhoravi]

As you can see all tests above have different PVs meaning one of the PVs maybe stronger than the other despite it took it longer to reach that depth.

How about deep endgame solving?