Upgrade to Chess.com Premium!

Ivom vs. markarkrkk

I light of recent comments made to my blog from Ivom, Ivom and I are now playing two correspondence games between our best engines.

I will play my Rybka 2.2n2 (32bit) against his Glaurung 2.2 (Linux OS), and give Toga II a chance to show its colours against Ivom's very own build of Rybka 2.2.

Games will be posted here once they have been played.

If you feel you can beat my Rybka 2.2 and Toga II engines hands down then please message me so we can do a similar thing (in an UNRATED game)

Comments


  • 3 years ago

    ivom

    I will just add a few comments about the Glaurung chess engine that I use in the game against Markarkrkk's Rybka.

    I started with a 32 bit version of linux and the latest version of Glaurung that comes with Ubuntu (Glaurung 2.2). Then I swithched to 64 bit version of Ubuntu, with the same version of Glaurung, but in a 64 bit mode.

    While I was impressed how Glaurung plays (it looks like it will beat Rybka in this particular game), the UCI adaptor polyglot, used to make the engine work with xboard was a bit slow in loading the engine. So I went and compiled the source code with "-O3" optimization instead of "-O2" recommended by the author, and used it instead of version that comes with Ubuntu. While I was at it I went and compiled to source code for Glaurung as well. As a result, when I pit Glaurung that comes with Ubuntu against the compiled version, with short time control of 5 minutes, the compiled version wins easily. 

    I also added Nalimov endgame tables, but I do not expect that to make any difference. First of all, I do not know if Glaurung uses them, and secondly, when the games comes down to 4 or 5 men the engine should be able to calculate well.

    I do not believe that any of the changes I made as the game progressed made any impact, and these are just some notes for those interested in using chess engines on Linux. I am still using the same engine, and given enough time, either 32 or 64 bit version should come up with the same move (unless there is some random number generator that would make a difference, I haven't looked at the actual source code).

Back to Top

Post your reply: