Hi Everyone,
EXchess versions 7.01/7.02 have an pondering bug related to the use of tablebases. I discovered it in online play this week. The fix was simple, but in the mean time I had made a few other changes that sped up the program, so I decided to include those in the bugfix as well.
The net result is EXchess v7.03 which is now available on my website...
http://personal.denison.edu/~homand/EXchess.htm
It should be a bit stronger (+10 elo perhaps) than v7.01/7.02 due to the improved speed. My apologies for producing a new version so quickly. I hope this addresses the last significant bug, and that 7.03 can be the last version for awhile.
Thanks!
- Dan
EXchess v7.03 released (bugfix + speed improved version)
Moderator: Ras
-
dchoman
- Posts: 177
- Joined: Wed Dec 28, 2011 8:44 pm
- Location: United States
- Full name: Daniel Homan
-
Graham Banks
- Posts: 45747
- Joined: Sun Feb 26, 2006 10:52 am
- Location: Auckland, NZ
Re: EXchess v7.03 released (bugfix + speed improved version)
A man's gotta do what a man's gotta do! 
Looking forward to trying it out on both 1 and 4 cores.
Can it use 8 cores as well?
Graham.
Looking forward to trying it out on both 1 and 4 cores.
Can it use 8 cores as well?
Graham.
gbanksnz at gmail.com
-
dchoman
- Posts: 177
- Joined: Wed Dec 28, 2011 8:44 pm
- Location: United States
- Full name: Daniel Homan
Re: EXchess v7.03 released (bugfix + speed improved version)
Thanks for understanding 
Yes, it can use up to 16 cores. To use more than 4, you will need to set the parameter LIMIT_THREADS in search.par to be larger than 4 (up to 16). This parameter sets the maximum allowed number of cores during play (the gui can then control the number from 1 up to this value).
- Dan
Yes, it can use up to 16 cores. To use more than 4, you will need to set the parameter LIMIT_THREADS in search.par to be larger than 4 (up to 16). This parameter sets the maximum allowed number of cores during play (the gui can then control the number from 1 up to this value).
- Dan
-
Graham Banks
- Posts: 45747
- Joined: Sun Feb 26, 2006 10:52 am
- Location: Auckland, NZ
Re: EXchess v7.03 released (bugfix + speed improved version)
Very good. Thanks.dchoman wrote:Thanks for understanding
Yes, it can use up to 16 cores. To use more than 4, you will need to set the parameter LIMIT_THREADS in search.par to be larger than 4 (up to 16). This parameter sets the maximum allowed number of cores during play (the gui can then control the number from 1 up to this value).
- Dan
gbanksnz at gmail.com
-
Max
- Posts: 247
- Joined: Tue Apr 13, 2010 10:41 am
Re: EXchess v7.03 released (bugfix + speed improved version)
Hi Daniel,
compiling (gcc 4.6) your engine on an ARM-based Linux system worked well. But Exchess 7.03 creates some strange mainlines here, while analyzing the start position. Anything to adjust for ARM architectures?
-Max
compiling (gcc 4.6) your engine on an ARM-based Linux system worked well. But Exchess 7.03 creates some strange mainlines here, while analyzing the start position. Anything to adjust for ARM architectures?
Code: Select all
$ ./exchess703
Experimental Chess Program (EXchess) version 7.03b (beta),
Copyright (C) 1997-2012 Daniel C. Homan, Granville OH, USA
EXchess comes with ABSOLUTELY NO WARRANTY. This is free
software, and you are welcome to redistribute it under
certain conditions. This program is distributed under the
GNU public license. See the files license.txt and readme.txt
for more information.
Hash size = 419430 buckets of 4 entries, 25 Mbytes
Pawn size = 62915 individual entries, 1 Mbytes
Score size = 220201 individual entries, 3 Mbytes
Cmove size = 52429 individual entries, 1 Mbytes
Type 'help' for a list of commands.
White-To-Move[1]: new
White-To-Move[1]: analyze
Analysis mode: Enter commands/moves as ready.
1. 2.28 0 2 1. a3
1. 3.06 0 4 1. b3
1. 3.06 0 21 1. b3
2. 0.13 0 43 1. b3 b6
2. 0.13 0 78 1. b3 b6
3. 3.06 0 161 1. b3 b6 2. g3
3. 3.06 0 190 1. b3 b6 2. g3
<-->
<-->
<-->
4. 0.13 0 402 1. b3 b6 2. g3 g6
4. 0.13 0 424 1. b3 b6 2. g3 g6
<-->
5. 0.95 0 524 1. b3 b6
6. 0.39 0 1057 1. b3 b6 2. g3 g5 3. h4 a5
6. 0.39 0 1098 1. b3 b6 2. g3 g5 3. h4 a5
7. 0.63 0 1426 1. b3 b6
<-->
8. 0.43 0 1834 1. b3 b6 2. g3 a6 3. h4 h6 4. a4 g6
8. 0.43 0 1878 1. b3 b6 2. g3 a6 3. h4 h6 4. a4 g6
9. 0.45 0 2091 1. b3 b6 2. g3 a6 3. h4 h6 4. a4 g6 5. c3
9. 0.45 0 2142 1. b3 b6 2. g3 a6 3. h4 h6 4. a4 g6 5. c3
10. 0.52 0 2389 1. b3 b6 2. g3 a6 3. h4 h6 4. a4 g6 5. c3 c5
10. 0.52 0 2440 1. b3 b6 2. g3 a6 3. h4 h6 4. a4 g6 5. c3 c5
11. 0.41 0 2712 1. b3 b6 2. g3 a6 3. h4 h6 4. a4 g6 5. c3 c5 6. b4
11. 0.41 0 2763 1. b3 b6 2. g3 a6 3. h4 h6 4. a4 g6 5. c3 c5 6. b4
<-->
12. 0.58 0 3268 1. b3 b6
13. 0.55 0 3828 1. b3 b6 2. g3 a6 3. h4 h6 4. a4 g6 5. c3 c6 6. b4 Qc7 7. Na3
13. 0.55 0 3879 1. b3 b6 2. g3 a6 3. h4 h6 4. a4 g6 5. c3 c6 6. b4 Qc7 7. Na3
14. 0.70 0 5808 1. b3 h6
<-->
15. 0.43 0 10831 1. b3 h6 2. g3 b6 3. h4 g6 4. a4 a5 5. c3 c6 6. b4 h5 7. Na3 axb4 8. cxb4
15. 0.43 0 10875 1. b3 h6 2. g3 b6 3. h4 g6 4. a4 a5 5. c3 c6 6. b4 h5 7. Na3 axb4 8. cxb4
16. 0.41 0 12590 1. b3 h6 2. g3 b6 3. h4 g6 4. a4 a5 5. c3 c6 6. b4 h5 7. Na3 axb4 8. cxb4 Na6
16. 0.41 0 12641 1. b3 h6 2. g3 b6 3. h4 g6 4. a4 a5 5. c3 c6 6. b4 h5 7. Na3 axb4 8. cxb4 Na6
17. 0.32 0 15427 1. b3 h6 2. g3 b6 3. h4 g6 4. a4 a5 5. c3 c6 6. b4 h5 7. Na3 axb4 8. cxb4 Na6 9. Nh3
17. 0.32 0 15478 1. b3 h6 2. g3 b6 3. h4 g6 4. a4 a5 5. c3 c6 6. b4 h5 7. Na3 axb4 8. cxb4 Na6 9. Nh3
18. 0.26 0 26494 1. b3 b6 2. g3 h6 3. h4
18. 0.26 0 26557 1. b3 b6 2. g3 h6 3. h4
<-->
19. 0.41 0 29765 1. b3 b6
20. 0.26 0 32497 1. b3 b6
<-->
21. 0.11 0 43222 1. b3 b6
22. -0.04 1 93986 1. b3 g6
<-->
23. 0.11 1 119465 1. b3 g6
24. -0.04 1 137672 1. b3 g6
<-->
25. 0.11 1 161723 1. b3 g6
26. 0.56 1 203519 1. b3 g6
<-->
27. 0.71 2 269391 1. b3 g6
28. 0.26 3 455564 1. b3 b6
<-->
29. 0.11 5 649116 1. b3 a5
30. 0.26 5 720080 1. b3 b6
<-->
31. 0.11 6 838212 1. b3 b6
32. 0.26 8 1131116 1. a3 g6
<-->
33. 0.11 10 1374115 1. a3 g6
34. -0.04 11 1535470 1. a3 h5
<-->
35. 0.11 12 1706952 1. a3 h5
36. -0.04 14 1945124 1. a3 h5
<-->
37. 0.11 17 2557318 1. a3 h5
38. 0.00 19 2861523 1. a3 h5 2. b4 b6 3. h4 g6 4. g4 a5 5. bxa5 bxa5 6. gxh5 gxh5 7. a4 Na6 8. c3 c6 9. Na3 Nh6 10. Nh3 f6 11. Bg2 Ng8 12. Bf1 Nh6
38. 0.00 21 3164513 1. a3 h5 2. b4 b6 3. h4 g6 4. g4 a5 5. bxa5 bxa5 6. gxh5 gxh5 7. a4 Na6 8. c3 c6 9. Na3 Nh6 10. Nh3 f6 11. Bg2 Ng8 12. Bf1 Nh6
39. 0.00 25 3775534 1. a3 h5 2. b4 a5 3. bxa5 g6 4. g4 b6 5. h4 bxa5 6. gxh5 gxh5 7. a4 c6 8. c3 Na6 9. Na3 Nh6 10. Nh3 f6 11. Bg2 Ng8 12. Bf1 Nh6
39. ++ 27 4031710 1. g3 g6
39. 0.15 28 4145429 1. g3 g6
<-->
40. 0.00 31 4635838 1. g3 g6
41. 0.15 41 6191016 1. g3 b6
quit
nodes = 6438912 hash moves = 272588 qnodes = 2035480 evals = 2235866
hash hits = 3058503 pawn hash hits = 1123207 score hash hits = 604707
node_rate = 146839 null cuts = 677585 exten = 23837 qchecks = 184
int_iter = 225027 egtb_probes = 0 egtb_hits = 0 fail_high(%) = 93
White-To-Move[1]: -
dchoman
- Posts: 177
- Joined: Wed Dec 28, 2011 8:44 pm
- Location: United States
- Full name: Daniel Homan
Re: EXchess v7.03 released (bugfix + speed improved version)
Those mainlines are indeed odd. I've never seen anything like that. It is like the piece-square tables are messed up in some fashion, but that is only a guess. I don't have an ARM architecture, so I cannot be sure....
One possibility, I suppose, is stack size.... but that is just a guess. You could try setting MAX_THREADS to 1 in define.h and recompile. If it works normally, then stack size was the problem, and you can either increase the stack size during the compile or just limit MAX_THREADS to the largest value you intend to use (and that works).
- Dan
One possibility, I suppose, is stack size.... but that is just a guess. You could try setting MAX_THREADS to 1 in define.h and recompile. If it works normally, then stack size was the problem, and you can either increase the stack size during the compile or just limit MAX_THREADS to the largest value you intend to use (and that works).
- Dan
-
Max
- Posts: 247
- Joined: Tue Apr 13, 2010 10:41 am
Re: EXchess v7.03 released (bugfix + speed improved version)
Sorry, but same behavior with this change.dchoman wrote:One possibility, I suppose, is stack size.... but that is just a guess. You could try setting MAX_THREADS to 1 in define.h and recompile
Perhaps there is a relationship to the issue I reported some time ago here?
-Max
-
dchoman
- Posts: 177
- Joined: Wed Dec 28, 2011 8:44 pm
- Location: United States
- Full name: Daniel Homan
Re: EXchess v7.03 released (bugfix + speed improved version)
Hummm.... OK, maybe it is a variable initialization problem? Something that on other systems is defaulting to a sensible value but not with this particular processor/compiler combination. My son has an old netbook with linux, but I am not sure what the processor is.... an intel Atom maybe? If I can reproduce the problem, I should be able to track it down. I'll work on it.Max wrote:Sorry, but same behavior with this change.dchoman wrote:One possibility, I suppose, is stack size.... but that is just a guess. You could try setting MAX_THREADS to 1 in define.h and recompile
Perhaps there is a relationship to the issue I reported some time ago here?
-Max
Thanks for pointing this out and reminding me of the old thread. I had completely forgotten about it... I'll compare v6.01 to v6.02 to see what changed, and that may also give a clue.
- Dan
-
Anton
- Posts: 3549
- Joined: Sun Mar 26, 2006 5:53 pm
Re: EXchess v7.03 released (bugfix + speed improved version)
Hi Daniel,
Great to see one of my old favorites is still around, tested the linux64 version with 1, 4, and 8 threads in console mode, everything worked just fine. Then compiled the source, gotta say that if you compiled the binary with g++ -o EXchess_console src/EXchess.cc -O2 -pthread, then I want your compiler, had to profile it here to get a binary as fast as the one you shipped, ran the fresh binary on 1, 4, and 8 threads, then fell asleep for a couple of hours with it still running, woke up to find everything still working.
Great to see one of my old favorites is still around, tested the linux64 version with 1, 4, and 8 threads in console mode, everything worked just fine. Then compiled the source, gotta say that if you compiled the binary with g++ -o EXchess_console src/EXchess.cc -O2 -pthread, then I want your compiler, had to profile it here to get a binary as fast as the one you shipped, ran the fresh binary on 1, 4, and 8 threads, then fell asleep for a couple of hours with it still running, woke up to find everything still working.
Code: Select all
Type 'help' for a list of commands.
White-To-Move[1]: hash 6300
Hash size = 82575360 buckets of 4 entries, 5040 Mbytes
Pawn size = 12386304 individual entries, 283 Mbytes
Score size = 43352064 individual entries, 661 Mbytes
Cmove size = 10321920 individual entries, 315 Mbytes
White-To-Move[1]: analyze
Analysis mode: Enter commands/moves as ready.
1. -0.28 0 15 1. a3
1. 0.33 0 66 1. b4
1. 0.74 0 95 1. d4
1. 1.08 0 109 1. e3
1. 1.08 0 109 1. e3
2. 0.13 0 221 1. e3 e6
2. 0.13 0 485 1. e3 e6
2. 0.13 0 493 1. e3 e6
3. 0.75 0 966 1. e3 e6 2. Nc3
3. 0.87 0 1489 1. e4 e6 2. d4
3. 0.87 0 1517 1. e4 e6 2. d4
3. 0.87 0 1517 1. e4 e6 2. d4
6. 0.25 0 17643 1. e4 Nc6 2. Nf3 e6 3. d4 Nf6
7. ++ 0 18372 1. e4 e6
7. 0.42 0 18374 1. e4 e6
8. 0.56 0 48510 1. e4 e5 2. Nc3 d6 3. Nf3 Nc6 4. d4 Nf6 5. Bb5
9. 0.57 0 64654 1. e4 Nc6
<-->
10. 0.37 0 235101 1. e4 d5 2. exd5 e6 3. Nc3 exd5 4. d4
10. 0.37 0 247773 1. e4 d5 2. exd5 e6 3. Nc3 exd5 4. d4 Nc6 5. Qe2+ Nge7 6. Nf3
10. 0.37 0 247775 1. e4 d5 2. exd5 e6 3. Nc3 exd5 4. d4 Nc6 5. Qe2+ Nge7 6. Nf3
11. 0.30 0 337790 1. e4 d5 2. exd5 e6
11. 0.30 0 360092 1. e4 d5 2. exd5 e6
11. 0.30 0 360097 1. e4 d5 2. exd5 e6
12. 0.39 0 442626 1. e4 d5 2. exd5
12. 0.39 0 442626 1. e4 d5 2. exd5
13. 0.50 0 850777 1. e4 e6 2. Nc3 d5
13. 0.50 0 916637 1. e4 e6 2. Nc3 d5
13. 0.50 0 916637 1. e4 e6 2. Nc3 d5
14. 0.51 0 1623286 1. e4 d5 2. exd5 Nf6 3. d4 e6 4. Bb5+
14. 0.51 0 1684560 1. e4 d5 2. exd5 Nf6 3. d4 e6 4. Bb5+
15. 0.36 0 2091148 1. e4 d5
16. ++ 0 2492094 1. e4
16. 0.51 0 2501594 1. e4 d5
<-->
17. 0.36 1 5712245 1. e4 e5
18. 0.39 1 7579204 1. e4 e5 2. Nf3 Nc6 3. Nc3 Nf6 4. d4 exd4 5. Nxd4 Bb4 6. Bg5 d6 7. Nxc6
18. 0.39 1 8107523 1. e4 e5 2. Nf3 Nc6 3. Nc3 Nf6 4. d4 exd4 5. Nxd4 Bb4 6. Bg5 d6 7. Nxc6
19. 0.24 2 11639916 1. e4 e5
<-->
20. 0.39 2 14240686 1. e4 e5
21. 0.24 3 21815258 1. e4 e5
<-->
22. 0.39 6 39030402 1. e4 e6
23. 0.24 14 85584678 1. e4 e5
24. 0.42 16 98894991 1. e4 e5 2. Nf3 Nc6 3. d4 exd4 4. Nxd4 Nf6
24. 0.42 16 99575841 1. e4 e5 2. Nf3 Nc6 3. d4 exd4 4. Nxd4 Nf6
<-->
25. ++ 25 147889613 1. e4 e5
25. 0.37 25 147889621 1. e4 e5
26. 0.49 34 199365611 1. e4 e5 2. Nf3 Nc6 3. d4
26. 0.49 35 207255089 1. e4 e5 2. Nf3 Nc6 3. d4
27. 0.49 38 226586086 1. e4 e5 2. Nf3
27. 0.49 40 233844083 1. e4 e5 2. Nf3
<-->
<-->
28. ++ 61 352857119 1. e4 e5
28. 0.34 61 352857119 1. e4 e5
<-->
29. 0.49 92 534095558 1. e4 e5
30. 0.34 141 809290202 1. e4 e5
31. ++ 180 1033995863 1. e4
31. 0.49 197 1128519244 1. e4 e5
<-->
32. 0.34 304 1720521574 1. e4 e5
33. 0.19 635 3544349928 1. e4 e5
34. ++ 695 3872235501 1. e4
34. 0.34 720 4006521443 1. e4 e5
<-->
<-->
35. 0.19 2169 11722866149 1. e4 e5
36. 0.34 2318 12548072165 1. e4 e5
<-->
37. 0.19 2855 15485917513 1. e4 e5
38. 0.34 3168 17201737988 1. e4 e5
<-->
40. ++ 6154 33533126746 1. e4 e5
40. 0.49 6435 35061667521 1. e4 e5
41. 0.37 7689 42028771989 1. e4 e5 2. Nf3 Nc6 3. d4
41. 0.37 8352 45693513090 1. e4 e5 2. Nf3 Nc6 3. d4
41. 0.37 8352 45693513090 1. e4 e5 2. Nf3 Nc6 3. d4
42. 0.37 8837 48407376016 1. e4 e5 2. Nf3
nodes = 50649212664 hash moves = 1481268318 qnodes = 28319173382 evals = 29316031619
hash hits = 5772852070 pawn hash hits = 23366234823 score hash hits = 5579674149
node_rate = 5487009 null cuts = 6769091962 exten = 1763875250 qchecks = 94439943
int_iter = 726597053 egtb_probes = 0 egtb_hits = 0 fail_high(%) = 89
-
dchoman
- Posts: 177
- Joined: Wed Dec 28, 2011 8:44 pm
- Location: United States
- Full name: Daniel Homan
Re: EXchess v7.03 released (bugfix + speed improved version)
Actually, I am now using -Ofast -flto -fwhole-program as my normal compile time options in g++ 4.6.3. I should probably update the readme.Anton wrote:Hi Daniel,
Great to see one of my old favorites is still around, tested the linux64 version with 1, 4, and 8 threads in console mode, everything worked just fine. Then compiled the source, gotta say that if you compiled the binary with g++ -o EXchess_console src/EXchess.cc -O2 -pthread, then I want your compiler, had to profile it here to get a binary as fast as the one you shipped, ran the fresh binary on 1, 4, and 8 threads, then fell asleep for a couple of hours with it still running, woke up to find everything still working.Code: Select all
Type 'help' for a list of commands. White-To-Move[1]: hash 6300 Hash size = 82575360 buckets of 4 entries, 5040 Mbytes Pawn size = 12386304 individual entries, 283 Mbytes Score size = 43352064 individual entries, 661 Mbytes Cmove size = 10321920 individual entries, 315 Mbytes White-To-Move[1]: analyze Analysis mode: Enter commands/moves as ready. 1. -0.28 0 15 1. a3 1. 0.33 0 66 1. b4 1. 0.74 0 95 1. d4 1. 1.08 0 109 1. e3 1. 1.08 0 109 1. e3 2. 0.13 0 221 1. e3 e6 2. 0.13 0 485 1. e3 e6 2. 0.13 0 493 1. e3 e6 3. 0.75 0 966 1. e3 e6 2. Nc3 3. 0.87 0 1489 1. e4 e6 2. d4 3. 0.87 0 1517 1. e4 e6 2. d4 3. 0.87 0 1517 1. e4 e6 2. d4 6. 0.25 0 17643 1. e4 Nc6 2. Nf3 e6 3. d4 Nf6 7. ++ 0 18372 1. e4 e6 7. 0.42 0 18374 1. e4 e6 8. 0.56 0 48510 1. e4 e5 2. Nc3 d6 3. Nf3 Nc6 4. d4 Nf6 5. Bb5 9. 0.57 0 64654 1. e4 Nc6 <--> 10. 0.37 0 235101 1. e4 d5 2. exd5 e6 3. Nc3 exd5 4. d4 10. 0.37 0 247773 1. e4 d5 2. exd5 e6 3. Nc3 exd5 4. d4 Nc6 5. Qe2+ Nge7 6. Nf3 10. 0.37 0 247775 1. e4 d5 2. exd5 e6 3. Nc3 exd5 4. d4 Nc6 5. Qe2+ Nge7 6. Nf3 11. 0.30 0 337790 1. e4 d5 2. exd5 e6 11. 0.30 0 360092 1. e4 d5 2. exd5 e6 11. 0.30 0 360097 1. e4 d5 2. exd5 e6 12. 0.39 0 442626 1. e4 d5 2. exd5 12. 0.39 0 442626 1. e4 d5 2. exd5 13. 0.50 0 850777 1. e4 e6 2. Nc3 d5 13. 0.50 0 916637 1. e4 e6 2. Nc3 d5 13. 0.50 0 916637 1. e4 e6 2. Nc3 d5 14. 0.51 0 1623286 1. e4 d5 2. exd5 Nf6 3. d4 e6 4. Bb5+ 14. 0.51 0 1684560 1. e4 d5 2. exd5 Nf6 3. d4 e6 4. Bb5+ 15. 0.36 0 2091148 1. e4 d5 16. ++ 0 2492094 1. e4 16. 0.51 0 2501594 1. e4 d5 <--> 17. 0.36 1 5712245 1. e4 e5 18. 0.39 1 7579204 1. e4 e5 2. Nf3 Nc6 3. Nc3 Nf6 4. d4 exd4 5. Nxd4 Bb4 6. Bg5 d6 7. Nxc6 18. 0.39 1 8107523 1. e4 e5 2. Nf3 Nc6 3. Nc3 Nf6 4. d4 exd4 5. Nxd4 Bb4 6. Bg5 d6 7. Nxc6 19. 0.24 2 11639916 1. e4 e5 <--> 20. 0.39 2 14240686 1. e4 e5 21. 0.24 3 21815258 1. e4 e5 <--> 22. 0.39 6 39030402 1. e4 e6 23. 0.24 14 85584678 1. e4 e5 24. 0.42 16 98894991 1. e4 e5 2. Nf3 Nc6 3. d4 exd4 4. Nxd4 Nf6 24. 0.42 16 99575841 1. e4 e5 2. Nf3 Nc6 3. d4 exd4 4. Nxd4 Nf6 <--> 25. ++ 25 147889613 1. e4 e5 25. 0.37 25 147889621 1. e4 e5 26. 0.49 34 199365611 1. e4 e5 2. Nf3 Nc6 3. d4 26. 0.49 35 207255089 1. e4 e5 2. Nf3 Nc6 3. d4 27. 0.49 38 226586086 1. e4 e5 2. Nf3 27. 0.49 40 233844083 1. e4 e5 2. Nf3 <--> <--> 28. ++ 61 352857119 1. e4 e5 28. 0.34 61 352857119 1. e4 e5 <--> 29. 0.49 92 534095558 1. e4 e5 30. 0.34 141 809290202 1. e4 e5 31. ++ 180 1033995863 1. e4 31. 0.49 197 1128519244 1. e4 e5 <--> 32. 0.34 304 1720521574 1. e4 e5 33. 0.19 635 3544349928 1. e4 e5 34. ++ 695 3872235501 1. e4 34. 0.34 720 4006521443 1. e4 e5 <--> <--> 35. 0.19 2169 11722866149 1. e4 e5 36. 0.34 2318 12548072165 1. e4 e5 <--> 37. 0.19 2855 15485917513 1. e4 e5 38. 0.34 3168 17201737988 1. e4 e5 <--> 40. ++ 6154 33533126746 1. e4 e5 40. 0.49 6435 35061667521 1. e4 e5 41. 0.37 7689 42028771989 1. e4 e5 2. Nf3 Nc6 3. d4 41. 0.37 8352 45693513090 1. e4 e5 2. Nf3 Nc6 3. d4 41. 0.37 8352 45693513090 1. e4 e5 2. Nf3 Nc6 3. d4 42. 0.37 8837 48407376016 1. e4 e5 2. Nf3 nodes = 50649212664 hash moves = 1481268318 qnodes = 28319173382 evals = 29316031619 hash hits = 5772852070 pawn hash hits = 23366234823 score hash hits = 5579674149 node_rate = 5487009 null cuts = 6769091962 exten = 1763875250 qchecks = 94439943 int_iter = 726597053 egtb_probes = 0 egtb_hits = 0 fail_high(%) = 89
Thanks for letting me know it worked so stably on your system for a multi-hour search!
- Dan