Perft(13) [3.4 GHz Core i7-2600, 16 GB RAM]

Discussion of chess software programming and technical issues.

Moderator: Ras

User avatar
Ajedrecista
Posts: 2141
Joined: Wed Jul 13, 2011 9:04 pm
Location: Madrid, Spain.

Re: Verification of large draft 8 record.

Post by Ajedrecista »

Hello Steven:
sje wrote:As mentioned earlier, some of the draft 8 record counts can be fairly large. Here's a verification run of one of the recent results done all by its lonesome self on a separate box:
[d]

Code: Select all

[] emptran 8
Bxg4 18,078,680,415
Bd7 74,010,050,049
Be6 62,470,891,901
Bf5 86,483,789,607
Na6 63,991,602,356
Nd7 47,745,655,289
Nh6 66,057,131,001
Nf6 77,039,439,053
Nc6 93,982,773,746
Qa5 85,744,207,928
Qc7 105,522,085,151
Qb6 106,367,651,283
Qd6 126,258,823,266
Qd7 72,422,310,827
a6 64,127,351,261
b5 68,807,933,023
a5 72,619,084,231
b6 59,822,349,656
c4 66,664,415,654
d4 73,968,079,707
f5 51,278,529,873
f6 54,475,865,708
e6 86,806,620,450
e5 104,787,610,258
g5 64,876,649,246
h6 62,529,998,177
g6 72,370,147,718
h5 77,925,261,293
Depth: 8   Count: 2,067,234,988,127   Elapsed: 490.107  (4.21793e+09 Hz / 2.37083e-10 s)
I can verify this draft 8 record:

[d]

Code: Select all

JetChess 1.0.0.0 using 1 GB of hash:

  1  qd8-c7  105522085151
  2  qd8-b6  106367651283
  3  qd8-a5   85744207928
  4  qd8-d7   72422310827
  5  qd8-d6  126258823266
  6  bc8-d7   74010050049
  7  bc8-e6   62470891901
  8  bc8-f5   86483789607
  9  bc8*g4   18078680415
 10  nb8-a6   63991602356
 11  nb8-c6   93982773746
 12  nb8-d7   47745655289
 13  ng8-f6   77039439053
 14  ng8-h6   66057131001
 15   c5-c4   66664415654
 16   d5-d4   73968079707
 17   a7-a6   64127351261
 18   a7-a5   72619084231
 19   b7-b6   59822349656
 20   b7-b5   68807933023
 21   e7-e6   86806620450
 22   e7-e5  104787610258
 23   f7-f6   54475865708
 24   f7-f5   51278529873
 25   g7-g6   72370147718
 26   g7-g5   64876649246
 27   h7-h6   62529998177
 28   h7-h5   77925261293

Total:      2067234988127

2,067,234,988,127 (move pathes after 8 half moves).

Time: 2482.723 s.
Please keep the good work!

Regards from Spain.

Ajedrecista.
Top
User avatar
sje
Posts: 4675
Joined: Mon Mar 13, 2006 7:43 pm

Try one of these

Post by sje »

Here are the ten most recent draft 9 records:

Code: Select all

rnbqkbnr/pp1pp1pp/8/2p2p2/8/N4P2/PPPPP1PP/R1BQKBNR w KQkq - 0 3 9 2606744811181
rnbqkbnr/pp1ppp1p/6p1/2p5/8/N4P2/PPPPP1PP/R1BQKBNR w KQkq - 0 3 9 2992519915722
rnbqkbnr/pp1ppp1p/8/2p3p1/8/N4P2/PPPPP1PP/R1BQKBNR w KQkq - 0 3 9 2943186301949
rnbqkbnr/pp1ppp1p/8/2p3p1/3P4/5N2/PPP1PPPP/RNBQKB1R w KQkq - 0 3 9 15058338794429
rnbqkbnr/pp1pppp1/7p/2p5/8/N4P2/PPPPP1PP/R1BQKBNR w KQkq - 0 3 9 2299084105993
rnbqkbnr/pp1pppp1/8/2p4p/8/N4P2/PPPPP1PP/R1BQKBNR w KQkq - 0 3 9 3107831024890
rnbqkbnr/pp1pp1pp/8/2p2p2/P2P4/8/1PP1PPPP/RNBQKBNR w KQkq - 0 3 9 13925721937336
r1bqkbnr/pp1ppppp/n7/2p5/8/N4P2/PPPPP1PP/R1BQKBNR w KQkq - 1 3 9 2502169364297
r1bqkbnr/pp1ppppp/2n5/2p5/8/N4P2/PPPPP1PP/R1BQKBNR w KQkq - 1 3 9 3999664320219
rnbqkbnr/pp1pp1pp/8/2p2p2/4P3/2N5/PPPP1PPP/R1BQKBNR w KQkq - 0 3 9 16447691087675
Here are the ten most recent draft 10 records:

Code: Select all

rnbqkbnr/pp1ppppp/8/2p5/8/N3P3/PPPP1PPP/R1BQKBNR b KQkq - 0 2 10 335154363673682
rnbqkbnr/pp1ppppp/8/2p5/3P4/P7/1PP1PPPP/RNBQKBNR b KQkq - 0 2 10 318173816347952
rnbqkbnr/pp1ppppp/8/2p5/8/3P1N2/PPP1PPPP/RNBQKB1R b KQkq - 0 2 10 295469329668234
rnbqkbnr/pp1ppppp/8/2p5/P7/3P4/1PP1PPPP/RNBQKBNR b KQkq - 0 2 10 246874945709599
rnbqkbnr/pp1ppppp/8/2p5/8/2N1P3/PPPP1PPP/R1BQKBNR b KQkq - 0 2 10 496400244445804
rnbqkbnr/pp1ppppp/8/2p5/3P4/1P6/P1P1PPPP/RNBQKBNR b KQkq - 0 2 10 363399058371809
rnbqkbnr/pp1ppppp/8/2p5/1P6/3P4/P1P1PPPP/RNBQKBNR b KQkq - 0 2 10 293974652424126
rnbqkbnr/pp1ppppp/8/2p5/3P4/7N/PPP1PPPP/RNBQKB1R b KQkq - 0 2 10 353688002077697
rnbqkbnr/pp1ppppp/8/2p5/8/P3P3/1PPP1PPP/RNBQKBNR b KQkq - 0 2 10 308960684953042
rnbqkbnr/pp1ppppp/8/2p5/4P3/N7/PPPP1PPP/R1BQKBNR b KQkq - 0 2 10 345274407616542
Top
ibid
Posts: 89
Joined: Mon Jun 13, 2011 12:09 pm

Re: Try one of these

Post by ibid »

sje wrote:Here are the ten most recent draft 10 records:

Code: Select all

rnbqkbnr/pp1ppppp/8/2p5/8/N3P3/PPPP1PPP/R1BQKBNR b KQkq - 0 2 10 335154363673682
rnbqkbnr/pp1ppppp/8/2p5/3P4/P7/1PP1PPPP/RNBQKBNR b KQkq - 0 2 10 318173816347952
rnbqkbnr/pp1ppppp/8/2p5/8/3P1N2/PPP1PPPP/RNBQKB1R b KQkq - 0 2 10 295469329668234
rnbqkbnr/pp1ppppp/8/2p5/P7/3P4/1PP1PPPP/RNBQKBNR b KQkq - 0 2 10 246874945709599
rnbqkbnr/pp1ppppp/8/2p5/8/2N1P3/PPPP1PPP/R1BQKBNR b KQkq - 0 2 10 496400244445804
rnbqkbnr/pp1ppppp/8/2p5/3P4/1P6/P1P1PPPP/RNBQKBNR b KQkq - 0 2 10 363399058371809
rnbqkbnr/pp1ppppp/8/2p5/1P6/3P4/P1P1PPPP/RNBQKBNR b KQkq - 0 2 10 293974652424126
rnbqkbnr/pp1ppppp/8/2p5/3P4/7N/PPP1PPPP/RNBQKB1R b KQkq - 0 2 10 353688002077697
rnbqkbnr/pp1ppppp/8/2p5/8/P3P3/1PPP1PPP/RNBQKBNR b KQkq - 0 2 10 308960684953042
rnbqkbnr/pp1ppppp/8/2p5/4P3/N7/PPPP1PPP/R1BQKBNR b KQkq - 0 2 10 345274407616542
Picked one at random to confirm...
[d]

Code: Select all

496400244445804 positions.  2101 seconds.  done.
-paul
Top
User avatar
sje
Posts: 4675
Joined: Mon Mar 13, 2006 7:43 pm

Re: Try one of these

Post by sje »

ibid wrote:Picked one at random to confirm...
[d]

Code: Select all

496400244445804 positions.  2101 seconds.  done.
That time is rather less than the average time it has taken my calculation for a draft 10 result: 2 hours 52 minutes. Then again, Symbolic is doing all sorts of calculations related to search and evaluation which have nothing to do with movepath enumeration.
Top
ibid
Posts: 89
Joined: Mon Jun 13, 2011 12:09 pm

Re: Try one of these

Post by ibid »

sje wrote:
ibid wrote:Picked one at random to confirm...
[d]

Code: Select all

496400244445804 positions.  2101 seconds.  done.
That time is rather less than the average time it has taken my calculation for a draft 10 result: 2 hours 52 minutes. Then again, Symbolic is doing all sorts of calculations related to search and evaluation which have nothing to do with movepath enumeration.
Actually we are roughly the same on this position -- the first part of the output had scrolled away so I couldn't post it. It took roughly two hours to compute the unique positions to 8 ply first, then 2101 seconds to compute that perft from there... sorry, definitely should have mentioned that.
Top
ibid
Posts: 89
Joined: Mon Jun 13, 2011 12:09 pm

Re: Try one of these

Post by ibid »

Another one with a bit more detail...

[d]

Code: Select all

$ ./uniq uniq 1
allocated 230000000 DBEntry's (7019 MB)
table size (512 MB)
22/1 positions.  0 seconds.
$ ./uniq uniq 2
allocated 230000000 DBEntry's (7019 MB)
table size (512 MB)
593/22 positions.  0 seconds.
$ ./uniq uniq 3
allocated 230000000 DBEntry's (7019 MB)
table size (512 MB)
8867/593 positions.  1 seconds.
$ ./uniq uniq 4
allocated 230000000 DBEntry's (7019 MB)
table size (512 MB)
139839/8867 positions.  0 seconds.
$ ./uniq uniq 5
allocated 230000000 DBEntry's (7019 MB)
table size (512 MB)
1685634/139839 positions.  2 seconds.
$ ./uniq uniq 6
allocated 230000000 DBEntry's (7019 MB)
table size (512 MB)
20358242/1685634 positions.  19 seconds.
$ ./uniq uniq 7
allocated 230000000 DBEntry's (7019 MB)
table size (512 MB)
211687507/20358242 positions.  274 seconds.
$ ./uniq uniq 8
Split #1 to #12:  211687507 pos,  544 sec.
allocated 230000000 DBEntry's (7019 MB)
table size (512 MB)
Reduce #1: 179947570 pos,  281 sec.
Reduce #2: 179969982 pos,  276 sec.
Reduce #3: 179918348 pos,  276 sec.
Reduce #4: 179936990 pos,  276 sec.
Reduce #5: 179966697 pos,  276 sec.
Reduce #6: 179942416 pos,  276 sec.
Reduce #7: 179951934 pos,  276 sec.
Reduce #8: 179943130 pos,  276 sec.
Reduce #9: 179947730 pos,  275 sec.
Reduce #10: 179943270 pos,  276 sec.
Reduce #11: 179933223 pos,  276 sec.
Reduce #12: 179922605 pos,  276 sec.
Merge #1 to #12:  2159323895 pos,  60772 max,  531 sec.
Complete:  4391 sec.
$ ./uniq mperft 10 8 6
thread  1 -- block   1 -- time     31 -- perft 884840040064
thread  0 -- block   0 -- time     32 -- perft 737960912880
thread  2 -- block   2 -- time     33 -- perft 4479645251483
thread  3 -- block   3 -- time     33 -- perft 2001533348009
thread  4 -- block   4 -- time     33 -- perft 2860943803212
thread  5 -- block   5 -- time     33 -- perft 1430365028293
thread  0 -- block   7 -- time     64 -- perft 2293592453038
thread  1 -- block   6 -- time     64 -- perft 2688132007039
[...]
thread  1 -- block 208 -- time   1226 -- perft 271997940963
thread  5 -- block 209 -- time   1228 -- perft 434810726543
thread  0 -- block 210 -- time   1245 -- perft 348178564041
thread  3 -- block 211 -- time   1251 -- perft 320767611082
thread  4 -- block 212 -- time   1252 -- perft 301713836305
thread  2 -- block 213 -- time   1258 -- perft 195194196277
thread  5 -- block 215 -- time   1258 -- perft 131866300581
thread  1 -- block 214 -- time   1260 -- perft 220244599650
246874945709599 positions.  1261 seconds.  done.
$
So about 1 hour 40 minutes for this one. The whole thing is not very automated -- starting with the fact that the FEN is hard-coded in. :)
Top
User avatar
sje
Posts: 4675
Joined: Mon Mar 13, 2006 7:43 pm

Draft 11 record number 41

Post by sje »

[d]
The perft(11) for the above: 3,105,144,188,541,894

41 (10.25%) down, 359 (89.75%) to go.
Top
User avatar
sje
Posts: 4675
Joined: Mon Mar 13, 2006 7:43 pm

Draft 11 record number 42

Post by sje »

[d]
The perft(11) for the above: 2,645,575,697,730,606
42 (10.5%) down, 358 (89.5%) to go.
Top
User avatar
sje
Posts: 4675
Joined: Mon Mar 13, 2006 7:43 pm

Draft 11 record number 43

Post by sje »

[d]
The perft(11) for the above: 4,357,323,324,696,016

43 (10.75%) down, 357 (89.25%) to go.

The above result is the largest count seen so far.
Top
User avatar
sje
Posts: 4675
Joined: Mon Mar 13, 2006 7:43 pm

Draft 11 record numbers 44 and 45

Post by sje »

[d]
The perft(11) for the above: 3,213,899,487,866,532

[d]
The perft(11) for the above: 3,593,594,503,531,601

45 (11.25%) down, 355 (88.75%) to go.

[edit] Estimated completion date: April 4th, 2012
Top

Return to “Computer Chess Club: Programming and Technical Discussions”