Perft(13), final fifth

[sje]

This is a continuation of thread http://www.talkchess.com/forum/viewtopic.php?t=42207

My current guess as to the progress of the perft(13) run is that it is about 4/5 of the way done and that it should complete in early August 2012, about one year after it was started. This is a bit longer than I had first estimated; it appears that the draft eight calculations with both sides advancing a center pawn in early ply take about four times as long as those which do not. This effect was much less with perft(12) and I'll guess that it will be much more with perft(14).

[sje]

Perft(11) for 1 d3 d5 is 14,915,052,041,878,031. This is the first of the sixteen "hard" draft 11 counts generated and is also the largest draft 11 count to date.

http://dl.dropbox.com/u/31633927/Perft/Perft13/draft11

[sje]

After 43 weeks:

The run has produced 679,913 draft 8 records, about 83% of the total needed. The average draft 8 record generation time is about 38.2 seconds.

There are now 243 draft 11 records, 60.75% of the 400 needed. The average draft 11 record generation time is about 29.7 hours.
http://dl.dropbox.com/u/31633927/Perft/Perft13/draft11

There are now 9 draft 12 records, 45% of the 20 needed. The average draft 12 record generation time is about 33.4 days.
http://dl.dropbox.com/u/31633927/Perft/Perft13/draft12

Uptime:

Code: Select all

23:44:28 up 3 days,  7:09,  2 users,  load average: 8.00, 8.01, 8.05

Recent output:

Code: Select all

r1bqkbnr/ppp1pppp/2np4/8/2B1P3/8/PPPPNPPP/RNBQK2R b KQkq - 3 3 8 733329865425
rnbqkbnr/pppp2pp/5p2/4p3/8/2P1P3/PPQP1PPP/RNB1KBNR b KQkq - 0 3 8 581051757872
rn1qkbnr/ppp1pppp/3p2Q1/8/3P4/7b/PPP1PPPP/RNB1KBNR b KQkq - 3 3 8 1279258513360
rn1qkbnr/ppp1pppp/8/3p4/6P1/4P2b/PPPP1P1P/RNBQKBNR b KQkq - 0 3 8 410077818553
rn1qkbnr/ppp1pppp/8/3p4/8/4P1Pb/PPPP1P1P/RNBQKBNR w KQkq - 1 3 9 21105877122954
rnbqkbnr/pppppp1p/8/8/5Pp1/2P5/PP1PPKPP/RNBQ1BNR b kq - 0 3 8 112369854986
rnbqkbnr/p1p1pppp/B2p4/1p6/8/1P2P3/P1PP1PPP/RNBQK1NR b KQkq - 0 3 8 451753954672
rnbqkbnr/pppppp1p/8/8/2P2Pp1/8/PP1PPKPP/RNBQ1BNR b kq - 0 3 8 128815856640
rnbqkbnr/pppp1ppp/8/4p3/6P1/3P4/PPPBPP1P/RN1QKBNR b KQkq - 0 3 8 620281156472
rnbqkbnr/pppp2pp/5p2/4p3/4P3/2P5/PPQP1PPP/RNB1KBNR b KQkq - 0 3 8 362333948203

[sje]

With 686,367 draft 8 records, the run is now in it's final sixth and has been recently averaging about 1,100 draft 8 records per day.

There have been a few electrical outages due to rain and the fact that the power lines in New Hampshire retain integrity mostly only in dry, sunny, and wind-free weather. Every power cycle causes a loss of all the draft 7 (and lesser) records.

My guess is that the next draft 12 record will be for 1 f4 as 14 of its 20 draft 11 dependents have completed. We'll see in a couple of weeks.

[sje]

Perft(11) for 1 e3 d6 is 16,408,433,376,099,590. This is the second of the sixteen "hard" draft 11 counts generated and is also the largest draft 11 count to date.

http://dl.dropbox.com/u/31633927/Perft/Perft13/draft11

[sje]

After 44 weeks:

The run has produced 687,924 draft 8 records, about 84% of the total needed. The average draft 8 record generation time is about 38.7 seconds.

There are now 247 draft 11 records, 61.75% of the 400 needed. The average draft 11 record generation time is about 29.9 hours.
http://dl.dropbox.com/u/31633927/Perft/Perft13/draft11

There are now 9 draft 12 records, 45% of the 20 needed. The average draft 12 record generation time is about 34.2 days.
http://dl.dropbox.com/u/31633927/Perft/Perft13/draft12

Uptime:

Code: Select all

00:06:18 up 1 day, 14:53,  2 users,  load average: 8.01, 8.02, 8.05

Recent output:

Code: Select all

r1bqkbnr/ppp1pppp/n2p4/8/3P4/7P/PPP1PPP1/RNBQKBNR w KQkq - 1 3 9 14081934703468
rnbqkb1r/pppp1ppp/4p2n/8/2P1P3/8/PP1PKPPP/RNBQ1BNR b kq - 2 3 8 515770067317
rnbqkbn1/ppppppp1/7r/7p/5P2/5N2/PPPPPKPP/RNBQ1B1R b q - 3 3 8 403455531904
rn1qkbnr/ppp1pppp/3p4/8/4P3/7b/PPPPBPP1/RNBQK1NR b KQkq - 1 3 8 637993106031
rnbqkb1r/ppppnppp/8/4p3/8/1QP5/PP1PPPPP/RNBK1BNR b kq - 3 3 8 373240936104
rnbqkb1r/pppp1ppp/4p2n/8/2P1P3/8/PP1PBPPP/RNBQK1NR b KQkq - 2 3 8 622977478094
r1bqkbnr/ppp1pppp/2np4/8/3P4/7P/PPP1PPPR/RNBQKBN1 b Qkq - 2 3 8 542817412234
rnbqkbn1/ppppppp1/7r/7p/5P2/2P5/PP1PPKPP/RNBQ1BNR b q - 0 3 8 283854785216
rnb1kbnr/pppp1ppp/4p3/6q1/5B2/3P4/PPPQPPPP/RN2KBNR b KQkq - 3 3 8 2557509120268
rnbqkb1r/pppp1ppp/4p2n/8/2P1P3/3B4/PP1P1PPP/RNBQK1NR b KQkq - 2 3 8 667697247928

[sje]

Perft(11) for 1 d4 d6 is 15,948,394,992,914,745. This is the third of the sixteen "hard" draft 11 counts generated so far.

[sje]

Perft(11) for 1 e4 d6 is 17,344,682,849,928,475. This is the fourth of the sixteen "hard" draft 11 counts generated and is also the largest draft 11 count to date.

[sje]

After 45 weeks:

The run has produced 695,725 draft 8 records, about 85% of the total needed. The average draft 8 record generation time is about 39.1 seconds.

There are now 252 draft 11 records, 63% of the 400 needed. The average draft 11 record generation time is about 30.0 hours.
http://dl.dropbox.com/u/31633927/Perft/Perft13/draft11

There are now 9 draft 12 records, 45% of the 20 needed. The average draft 12 record generation time is about 35.0 days.
http://dl.dropbox.com/u/31633927/Perft/Perft13/draft12

Uptime:

Code: Select all

3 days, 15:03

Recent output:

Code: Select all

rnb1kbnr/ppppqppp/4p3/8/3P4/4B1P1/PPP1PP1P/RN1QKBNR b KQkq - 2 3 8 618724289363
r1bqkbnr/pppp1ppp/2n5/4p3/2P1Q3/8/PP1PPPPP/RNB1KBNR b KQkq - 3 3 8 1108657400375
r1bqkb1r/pppppppp/2n2n2/8/5P1P/4P3/PPPP2P1/RNBQKBNR b KQkq - 0 3 8 479690575613
rnbqkbnr/pppp1p1p/6p1/4p1B1/3P4/2P5/PP2PPPP/RN1QKBNR b KQkq - 1 3 8 1005832022641
r1bqkb1r/pppppppp/2n2n2/5P2/8/4P3/PPPP2PP/RNBQKBNR b KQkq - 0 3 8 382035732718
r1bqkb1r/pppppppp/2n2n2/8/5P2/4P3/PPPP2PP/RNBQKBNR w KQkq - 1 3 9 14189342447405
rnb1kbnr/ppppqppp/4p3/8/3P1B2/6P1/PPP1PP1P/RN1QKBNR b KQkq - 2 3 8 913077156212
rnbqkb1r/ppp1pppp/5n2/3p4/8/4PQ2/PPPPBPPP/RNB1K1NR b KQkq - 3 3 8 1352987840006
rnbqkbnr/ppp1pp1p/8/3p2p1/3P1B2/7P/PPP1PPP1/RN1QKBNR b KQkq - 0 3 8 752182027075
rnbqkbnr/ppp1pppp/8/8/4p3/5Q2/PPPPBPPP/RNB1K1NR b KQkq - 1 3 8 2168748745594

[sje]

The 1 f4 worker thread has completed.
[d]
The perft(12) for the above is 61,027,789,644,649,681.

This is the tenth draft 12 result so far; there are ten more to go.

The worker thread for 1 g4 has started. My estimate is that the next worker thread to finish will be the 1 c3 thread and this will take perhaps two more weeks.