World Draughts Forum

No posts since 2002?

PERFORMING NODE COUNT THROUGH DEPTH 18 ================================================================= --------------------------------- Node Count Through 18 Plies Completed |---| w |---| w |---| |---| | --------------------------------- Nodes from move # 1 = 550829166472 | w |---| w |---| |---|...

It occurred to me that: 1. perft is an unusual name for a function that performs a node count. It comes from Dr. Hyatt's Cray Blitz chess code (later Crafty) which he called "Performance Test", and, as usual back in those days, the name was truncated for use on a text-based programmatic interface. 2...

What do you mean exactly by "6-piece perfect play databases" ? I guess it is the optimisation of the distance to win but I am not familiar with checkers rules. Hi, Yes, I solved "Distance to Win" for the 6-piece database (and larger databases) for checkers. This is different from your "Distance to ...

By the way, I gave Jon Kreuzer, the GUI Checkers author, my 6-piece perfect play databases and the access code to probe them in RAM (over a year ago actually).

He says it plays much better with them.

I see he has not released it yet, I sent him an email inquiring about it.

void perft_per_ply(unsigned short depth, unsigned short white_to_move) { unsigned short number_of_legal_moves, i; unsigned long long total_nodes = 0; unsigned long long nodes_per_ply[MAX_LEGAL_MOVES_POSSIBLE_PER_POSITION]; CHECKERS_POSITION array_of_legal_moves[MAX_LEGAL_MOVES_POSSIBLE_PER_POSITION...

and now 2,041,959,240,377 agrees with yours as well.

Here are my perft with divide numbers that I recently re-computed to make sure they matched Aart's numbers. I just made this subject line so Aart would raise his eyebrows the next time he logs in, I did not computer perft(24) :) I can appreciate that sense of humor. I almost started my perft(25) co...

I have confirmed this number for perft(19):

12-16: 2,517,202,147,314

Still doing the other 6 now

Well, I have a 3.9 GHz Intel Core i7-860 x 4 processors with 8 GB of RAM. If I run your program with a much larger hash, say 2 GB or 4GB, I should get an even bigger performance increase (more hash entries) so it should even be faster, correct? Email it to me and I will run it. I would be curious to...

But I WAS using my small board diagrams

My really big ones are online for my database generator.

perft(16) 165629569428 nodes, 149.60s, 1107.15 Mnps perft(17) 758818810990 nodes, 522.30s, 1452.84 Mnps perft(18) 3493881706141 nodes, 1851.01s, 1887.55 Mnps perft(19) 16114043592799 nodes, 6465.48s, 2492.32 Mnps perft(20) 74545030871553 nodes, 22457.55s, 3319.38 Mnps perft(21) 345100524480819 node...

Hello everyone, Here are my perft with divide numbers that I recently re-computed to make sure they matched Aart's numbers. I just made this subject line so Aart would raise his eyebrows the next time he logs in, I did not computer perft(24) :) I also started at depth 16 since there was "too much te...

BertTuyt wrote: What I now do is to open ( fopen() )

I also remember reading that "fopen()" is now one of those "old-time" function calls, and it has been made "obsolete" but still able to be accessed.

I forget what the name of the new routine is, I think fopen64(). Same goes for fclose() and fseek(), etc.

Im now confronted with a limitation of opened files in Windows. Apparently this limit is 512 and can be set to 2048, but thats the upper maximum. Bert This sounds familiar. I only use a maximum of 400 files for all 10-piece database slices in 8x8 checkers, but I encountered it elsewhere on another ...