Thanks for the tests Werner. I guess it's just a bit weaker, which would make its result in the WCRCC that much more surprising. I'm also pretty stumped about the SMP benchmarks you posted in the WB forum. I'm looking into these issues, but I have very little time these days. I'll also try to get the use-cpu-time-when-idle thing fixed.
I'm getting more and more eager to rewrite the eval. I think it's probably the only thing holding ZCT back from a much higher rating.
Zach Wegner wrote:Thanks for the tests Werner. I guess it's just a bit weaker, which would make its result in the WCRCC that much more surprising. I'm also pretty stumped about the SMP benchmarks you posted in the WB forum. I'm looking into these issues, but I have very little time these days. I'll also try to get the use-cpu-time-when-idle thing fixed.
I'm getting more and more eager to rewrite the eval. I think it's probably the only thing holding ZCT back from a much higher rating.
Zach
Hi Zach,
perhaps there is a difference between wb and uci implementation?
I am running my match with the wb engine at the moment.
Console test was also under winboard.
I now test smp/single version under uci...