Hi Petr, What I mean is that games are simulated in parallel, but are one clock cycle apart from each other.
On Sun, Jun 20, 2010 at 8:51 AM, Petr Baudis <[email protected]> wrote: > Hi! > > On Thu, Jun 17, 2010 at 10:04:58PM +0800, Fuming Wang wrote: > > > Do all the games have to start synchonously? > > > > > > > > Games are started one clock at a time. > > I'm sorry, I don't quite understand that - does that mean that > different games can start at different times? > > > I do not quite understand your description. I'm guessing that you mean > to > > have n playouts at a time instead of one playout at a time like the > regular > > UCT algorithm. Nonetheless, FPGA based method are very static in > structure, > > and can not adapt easily to the dynamic changes of a game tree like a > > software implemented tree structure. So I don't know how to implement UCT > > type dynamic tree search algorithms in FPGA right now. > > I haven't meant the tree search to be done on the FPGA at all, I have > merely tried to draft a method to effeciently queue games to be played > on the FPGA next. Are you planning to do the alpha-beta search in the > FPGA? > > I see now. You were thinking about sending games to the FPGA and using the simulation results in the main computer. This is not what we are doing right now, we are doing the tree search in the FPGA. Fuming
_______________________________________________ Computer-go mailing list [email protected] http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
