Search found 3963 matches

by sje
Sun Aug 07, 2016 4:32 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Perft(14) Weekly Status Reports for 2016
Replies: 75
Views: 23858

Perft(14) Weekly Status 2016-08-07

Perft(14) Weekly Status 2016-08-07 Symbolic has produced more than 53,600,000 perft(7) results so far, about 55.60% of the 96,400,068 needed. Day count: 732 Estimated remaining day count: 585 Estimated total day count: 1,317 Average throughput: 73,224 results/day Effective frequency: 226.00 GHz Work...
by sje
Sat Aug 06, 2016 11:58 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Using side to move as a separate bit in hash key
Replies: 11
Views: 2658

Re: Splitting the table

Another idea is to split the table into two equal size tables; one for WTM, the other for BTM. I've done this and it has worked well. In a program with a variable search depth, both tables see about the same usage regardless of the root STM. Pre-determining which half of the hash table will be used...
by sje
Sat Aug 06, 2016 9:28 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Using side to move as a separate bit in hash key
Replies: 11
Views: 2658

Re: Splitting the table

Another idea is to split the table into two equal size tables; one for WTM, the other for BTM. I've done this and it has worked well. In a program with a variable search depth, both tables see about the same usage regardless of the root STM. Pre-determining which half of the hash table will be used...
by sje
Sat Aug 06, 2016 5:48 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Using side to move as a separate bit in hash key
Replies: 11
Views: 2658

Re: Splitting the table

Another idea is to split the table into two equal size tables; one for WTM, the other for BTM. I've done this and it has worked well. In a program with a variable search depth, both tables see about the same usage regardless of the root STM. Pre-determining which half of the hash table will be used...
by sje
Sat Aug 06, 2016 8:44 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Using side to move as a separate bit in hash key
Replies: 11
Views: 2658

Splitting the table

Another idea is to split the table into two equal size tables; one for WTM, the other for BTM. I've done this and it has worked well. In a program with a variable search depth, both tables see about the same usage regardless of the root STM. Bonus #0: No need for an extra XOR operation for position ...
by sje
Sat Aug 06, 2016 8:30 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Six perft(7) mismatch positions from work unit 533
Replies: 2
Views: 825

Re: Six perft(7) mismatch positions from work unit 533

Thank you for your verification work.

----

Symbolic has produced more than 53,600,000 perft(7) results so far, about 55.60% of the 96,400,068 needed.

That's just past the 5/9ths mark.

The effective rate has reached another high: 226 GHz
by sje
Sat Aug 06, 2016 5:11 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Six perft(7) mismatch positions from work unit 533
Replies: 2
Views: 825

Six perft(7) mismatch positions from work unit 533

Six perft(7) mismatch positions from work unit 533: 21194c21194 < rnbqkbnr/2p1pppp/p2p4/1p6/6P1/4PK2/PPPP1P1P/RNBQ1BNR b kq - 0 4 24 8915550431 213973210344 --- > rnbqkbnr/2p1pppp/p2p4/1p6/6P1/4PK2/PPPP1P1P/RNBQ1BNR b kq - 0 4 24 8915539463 213972947112 21596c21596 < rnbqkbnr/2p1pppp/p2p4/1p6/8/2P1P...
by sje
Sun Jul 31, 2016 8:44 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Perft(14) Weekly Status Reports for 2016
Replies: 75
Views: 23858

Perft(14) Weekly Status 2016-07-31

Perft(14) Weekly Status 2016-07-31 Symbolic has produced more than 52,700,000 perft(7) results so far, about 54.67% of the 96,400,068 needed. Day count: 725 Estimated remaining day count: 601 Estimated total day count: 1,326 Average throughput: 72,690 results/day Effective frequency: 221.93 GHz Work...
by sje
Fri Jul 29, 2016 8:02 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Location of files for the Perft(14) project
Replies: 0
Views: 608

Location of files for the Perft(14) project

Location of files for the Perft(14) project There are some challenges affecting the continuation of the Perft(14) project, so here are the public locations of the files needed for independent continuation if such becomes needed. The work is more than 50% complete, so it's all downhill from here. ---...
by sje
Thu Jul 28, 2016 12:44 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Twenty-nine perft(7) mismatches from work unit 528
Replies: 4
Views: 1374

Re: Twenty-nine perft(7) mismatches from work unit 528

Thank you for your verification calculations.

----

Symbolic has produced more than 52,700,000 perft(7) results so far, about 54.67% of the 96,400,068 needed.