No, he didn't.Tony Thomas wrote:Why is it that the guy who posted the results above has almost the same results as kat Kuhaut Ko?? Did he start a new account?
I am a different person actually, guess the forum software stores the user IPs, so the admin could verify that I am posting from Germany and not from Singapore.
I have been a passive reader of this engine experts forum for a while and thought I should share my observations regarding the rybka hash size when I came across this thread.
Thanks for your feedback, and for the power of 2 hint, did not realized that before.
For my CPU is rather old and slow, I usually try to find out the optimal settings for my programs, so with rybka. Unfortunately rybka does not tell anything about its hash usage (percentage off filling) like other engines, so one has to try and find out himself what is best.
To me it looks like there is some influence the way rybka works internally due to different hash size. And it may not always follow the rule: larger = better.
Perhaps they also use different algorithms depending on hash size. An indication could be that with 64 MB hash rybka doesn't go further than depth 10 in my testing position
(FEN: r3r3/p2n4/2qpp2k/1p5n/3PNp1p/1PP1QP2/P1B2P2/6RK w - - 0 1)
even after 1 hour, probably due to a bug which is unveiled only with a 64 MB hash algorithm?
Another indication could be the time to find the winning move which is 3-4 times faster with 64 MB hash.
Think this needs further investigation, as proposed by Ernest I will do some additional tests with a couple of different positions at the weekend.