Search found 3819 matches

by Henk
Thu Jul 04, 2013 10:02 am
Forum: Computer Chess Club: General Topics
Topic: back to the Komodo SMP issue
Replies: 79
Views: 16196

Re: back to the Komodo SMP issue

So probably at search levels above 18 the average branching factor should be greater than 2 otherwise searching faster doesn't give better games in terms of ELO rating.
by Henk
Thu Jul 04, 2013 7:19 am
Forum: Computer Chess Club: General Topics
Topic: back to the Komodo SMP issue
Replies: 79
Views: 16196

Re: back to the Komodo SMP issue

I computed 1->4 cores times-to-depth for several top engines, which on rating lists have comparable Elo gain from 1 to 4 cores. 150 positions, 30 positions repeated 5 times. Each run lasts one-two hours. Houdini 3 3.05 Stockfish 3 2.91 Rybka 4.1 2.32 Komodo 5.1 1.68 Crafty 23.5? 3.20 (computed ...
by Henk
Thu Jul 04, 2013 7:11 am
Forum: Computer Chess Club: General Topics
Topic: back to the Komodo SMP issue
Replies: 79
Views: 16196

Re: back to the Komodo SMP issue

I computed 1->4 cores times-to-depth for several top engines, which on rating lists have comparable Elo gain from 1 to 4 cores. 150 positions, 30 positions repeated 5 times. Each run lasts one-two hours. Houdini 3 3.05 Stockfish 3 2.91 Rybka 4.1 2.32 Komodo 5.1 1.68 Crafty 23.5? 3.20 (computed ...
by Henk
Wed Jul 03, 2013 6:18 am
Forum: Computer Chess Club: General Topics
Topic: Borislav Ivanov: a Lilov's add-on
Replies: 336
Views: 56235

Re: Receiving moves // Re: Borislav Ivanov: a Lilov's add-on

Bionic eye seems to be a bit too advanced and still out of reach for average people. This invisible headphone is quite possible http://arstechnica.com/gadgets/2013/06/man-creates-invisible-headphones-by-implanting-magnets-into-his-ears/ If one of the audience who is watching sends the move thru sms...
by Henk
Tue Jul 02, 2013 9:06 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: The simplicity cliche
Replies: 27
Views: 4049

Re: The simplicity cliche

Chess programming is all about performance. Simplicity, clarity might be a way to achieve this. Performance is crucial here. Or are we writing quasi mathematical poems. Possible too. Sorry, I thought we talk about programming in general. At least this is what I understood from the topic here. Maybe...
by Henk
Tue Jul 02, 2013 6:06 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: The simplicity cliche
Replies: 27
Views: 4049

Re: The simplicity cliche

[quote="nionita"][quote="bob"][quote="nionita"][quote="bob"][quote="nionita"]But complex things can be combined from simpler ones (composition). When we will have even better optimizing compilers, this will be the way.[/quote] Problem is, I am 65 years old now. I won't be programming long enough to ...
by Henk
Mon Jul 01, 2013 5:07 pm
Forum: Computer Chess Club: General Topics
Topic: Tarrasch GUI 2.02 - Ippolit 2 - miniMardi for Android etc.
Replies: 3
Views: 843

Re: Tarrasch GUI 2.02 - Ippolit 2 - miniMardi for Android et

Hello, here are some new released editions: http://www.triplehappy.com/ http://www.mediafire.com/download/8g33fpify36p9ah/Ippolit2_091.zip http://www.aartbik.com/MISC/eng.html http://sourceforge.net/projects/schach/ http://sourceforge.net/projects/wowchess/ Unfortunately I have no access to Ron´s C...
by Henk
Mon Jul 01, 2013 3:49 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hashing based on move lists
Replies: 108
Views: 11793

Re: Hashing based on move lists

The main point of this is that it has a much lower collision rate than Zobrist keys, even 128-bit ones, but 128 bits is impractical for anything other than Perft, IMO. On what do you base that conclusion :shock: I can already see a lot of trivial ways to make your keys collide... Please demonstrate...
by Henk
Mon Jul 01, 2013 8:26 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hashing based on move lists
Replies: 108
Views: 11793

Re: Hashing based on move lists

But such a gain is very unlikely to be found by someone who obviously doesn't have the slightest idea what he is doing, dreaming up stuff behind his desk without making even the simplest tests on it, so that even the most disastrously flawed schemes make it up to these pages, and still are defended...
by Henk
Sun Jun 30, 2013 4:19 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hashing based on move lists
Replies: 108
Views: 11793

Re: Hashing based on move lists

In my chess program, which has a dreadful performance by the way, I use the hashcode to look up an entry. If there are more entries with the same hashcode an equality operation is applied on the board representation. So you store the whole board representation in the hash entry. That is not very co...