OpenChess

Has anyone put Remi's new optimization tool to use on a cluster yet? Bob? If so, what kind of improvements are you seeing?

Has anyone put Remi's new optimization tool to use on a cluster yet? Bob? If so, what kind of improvements are you seeing?

On a cluster with N identical computers testing time is divided by N.

Michel Van den Bergh wrote:

Has anyone put Remi's new optimization tool to use on a cluster yet? Bob? If so, what kind of improvements are you seeing?

On a cluster with N identical computers testing time is divided by N.

Thanks for the input. I'd be very curios to see how CLOP performs versus Bob's manual single-variable tuning in terms of empirical data. Want to help me push Bob off his rocker?

gaard wrote:

Michel Van den Bergh wrote:

Has anyone put Remi's new optimization tool to use on a cluster yet? Bob? If so, what kind of improvements are you seeing?

On a cluster with N identical computers testing time is divided by N.

Thanks for the input. I'd be very curios to see how CLOP performs versus Bob's manual single-variable tuning in terms of empirical data. Want to help me push Bob off his rocker?

I have not tried it. Last time I spent a ton of time with simulated annealing, and the net result was zero. Too many terms involved... If I find time, I may give it a try at some point, but I won't expect much from it from past experience...