Winboard Forum

I've only posted to this board a few times, but the help here has enabled me to write up a bitboard engine successfully.

I was wondering more about performance metrics:

- if I run a 32-bit program on a 32 bit OS (like WinXP) with a 64 bit processor, do Bitboards lose their effectiveness (I assume so)

- are there any journal-type papers which discuss so much of the stuff you discuss here, as the concept may be tangentially related to some PhD work I'm doing, and I'm having a devil of a job pinning down sources to convince my supervisors of the performance benefits.

Cheers,

Yes, when you compile a bitboard engine for 32 bit, it is slower then when compiled for 64 bit. For my engine, the overall slowdown is a factor of 2; for all other engines, smaller factors were reported.

AFAIR, Tord found bitboards to be slower then mailbox on Glaurung 1 when compiled for 32 bit. I have no mailbox implementation of my engine, but I strongly assume that its performance would be between the 32 bit and the 64 bit compile of the bitboard implementation.

I don't think that the stuff that is discussed here is published in scientific papers. Nobody here seems to want to do that, while sometimes much shallower papers pass the reviews. See
http://64.68.157.89/forum/viewtopic.php?t=20913
for a discussion.

What's the subject of your PhD thesis?

John Digsby wrote:I've only posted to this board a few times, but the help here has enabled me to write up a bitboard engine successfully.

I was wondering more about performance metrics:

- if I run a 32-bit program on a 32 bit OS (like WinXP) with a 64 bit processor, do Bitboards lose their effectiveness (I assume so)

- are there any journal-type papers which discuss so much of the stuff you discuss here, as the concept may be tangentially related to some PhD work I'm doing, and I'm having a devil of a job pinning down sources to convince my supervisors of the performance benefits.

Cheers,

I am a bit unsure about mentioning the chess programming wiki here, because of former long years host and initiator of this forum, Volker Pittlik's harsh reaction on adertising the cpw some months ago. Anyway, you'll found some links on papers (as well as bitboard stuff) there, for instance the icga journal.
http://www.cs.unimaas.nl/icga/journal/
http://www.cs.unimaas.nl/icga/journal/c ... caj_db.pdf
where your university bib may have issues from.

Cheers,
Gerd

http://chessprogramming.wikispaces.com/

Gerd Isenberg wrote:

John Digsby wrote:I've only posted to this board a few times, but the help here has enabled me to write up a bitboard engine successfully.

I was wondering more about performance metrics:

- if I run a 32-bit program on a 32 bit OS (like WinXP) with a 64 bit processor, do Bitboards lose their effectiveness (I assume so)

- are there any journal-type papers which discuss so much of the stuff you discuss here, as the concept may be tangentially related to some PhD work I'm doing, and I'm having a devil of a job pinning down sources to convince my supervisors of the performance benefits.

Cheers,

I am a bit unsure about mentioning the chess programming wiki here, because of former long years host and initiator of this forum, Volker Pittlik's harsh reaction on adertising the cpw some months ago. Anyway, you'll found some links on papers (as well as bitboard stuff) there, for instance the icga journal.
http://www.cs.unimaas.nl/icga/journal/
http://www.cs.unimaas.nl/icga/journal/c ... caj_db.pdf
where your university bib may have issues from.

Cheers,
Gerd

http://chessprogramming.wikispaces.com/

Hi Gerd

Fear nothing

I'll add this useful resource when I update the mainpage links here.

Olivier

Thanks, Olivier!

Have realised that I forgot to reply to this thread.

I was interested (and still am) in using the concept of bitboards in a non-gaming field in order to perform rapid operations on locations. So far, the theory looks sounds, and is a lot easier than programming chess because there are so many fewer rules!

In my private programming, I have successfully implemented a bitboard engine for chess, and am very pleased with the results. Soons I will begin work on the AI, which is a little more complex...

Many thanks to all for the resources you provide/link to, which have inspired some of my current work.

Best wishes,
John