On Fri, Jun 11, 2010 at 8:52 PM, Mark Boon <[email protected]> wrote: > On Thu, Jun 10, 2010 at 11:09 PM, Erik van der Werf > <[email protected]> wrote: >> Lukasz Lew's libego used to have the fastest light playouts. IIRC some >> years ago it already got over 100k playouts per second for 9x9 on one >> core. I guess it should now be possible to get over 200k light >> playouts per second. Others on this list may have more accurate >> numbers. > > Cores didn't really get much faster recently, so what makes you think > it could be twice as fast now?
I think I remember seeing some benchmarks suggesting that the latest i7's should give roughly twice of what I was getting on a Core2 from some 3 years ago. I don't think they even had to overclock (which apparently is quite easy as well)... Anyway, I could be wrong, or it could depend on how you count logical cores. Better if someone replies who actually has new hardware, but otherwise just assume something between 100k and 200k (for pure light playouts, no tree). > The other thing that needs care when touting playout speeds is whether > this is purely measuring playouts, or playouts while building a MCTS > tree (or hashtable). Right, but when the tree comes in you really can't talk about a constant playout speed any more. As the tree grows you will spend more time in the tree and less in the playouts. > My Java implementation of light playouts with liberties does something > like 30K-35K per second on one core. I think it's similar on a 2.6Ghz > i5 as on an older 2.8Ghz. The same code directly translated to C is > about twice as fast. The strongest programs have the slowest playouts :-) Erik _______________________________________________ Computer-go mailing list [email protected] http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
