detects 2 CPUs on QX6700 and I7 920mcostalba wrote:Could you (and also all the other people that experience problems with CPU detection) please post type of CPU you have ? Thanks.
It's actually pretty similar to the way I do it in Daydreamer (at depth 32 I would reduce by 10.5-12 depending on the current evaluation and search bounds, which is probably a bit crazy):Dann Corbit wrote:it's a new and very interesting idea (I never saw it before) that will prune more and more aggressively as a function of existing search depth.Aaron Becker wrote:Here is the new nullmove depth reduction calculation:nepossiver wrote:Was Dann Corbit smooth scaling introduced into the source? I remember the few tests done showed either no difference or a 20-30 increase in ELO (which is what I found in 1+1 games).
I don't know how close this is to Dann's formulation.Code: Select all
// Null move dynamic reduction based on depth int R = 3 + (depth >= 5 * OnePly ? depth / 8 : 0); // Null move dynamic reduction based on value if (refinedValue - beta > PawnValueMidgame) R++;
For instance, if you are at a depth of 32, R would be 3+4=7.
Did you ever see a 7 ply reduction before? It's the most unheard of thing I have ever heard of!
Code: Select all
float null_r = 2.0 + ((depth + 2.0)/4.0) +
CLAMP(0, 1.5, (lazy_score-beta)/100.0);
Code: Select all
printf("%d %d %d\n", nLogicalCPU, nCores, builtin_cpu_count());Code: Select all
int cpu_count() {
return HT_enabled() ? builtin_cpu_count() / 2 : builtin_cpu_count();What is the philosophical idea behind auto detecting then number of cores? Based on the things I read before, in the spirit of SF it looked like it should leave this task to the interface. Right?mcostalba wrote:Could you (and also all the other people that experience problems with CPU detection) please post type of CPU you have ? Thanks.alpha123 wrote: I'd agree with you. It detected 2 threads on my quad, fortunately I saw that before I started testing it on playchess. It does use all 4 when set too, though.
Guys, maybe a 1.7.1 with proper thread detection and null move bug fix is needed.
Peter
The null move bug will require regression test anyway so although the fix could be quick the testing will take some time and we would wait at least a week or two to collect all the bug reports before to release a mainteinance version.
Intel Core 2 Quad Q8300 @ 2.50gHz, 6GB DDR2, Windows 7 Home Premium 64-bit.mcostalba wrote:Could you (and also all the other people that experience problems with CPU detection) please post type of CPU you have ? Thanks.alpha123 wrote: I'd agree with you. It detected 2 threads on my quad, fortunately I saw that before I started testing it on playchess
Guys, maybe a 1.7.1 with proper thread detection and null move bug fix is needed.
Peter
The null move bug will require regression test anyway so although the fix could be quick the testing will take some time and we would wait at least a week or two to collect all the bug reports before to release a mainteinance version.
I think the idea is that you can set the number of threads to use by the program as an UCI option, but the program then still has to see if there are enough cores available or else it will severely underperform.michiguel wrote:What is the philosophical idea behind auto detecting then number of cores? Based on the things I read before, in the spirit of SF it looked like it should leave this task to the interface. Right?mcostalba wrote:Could you (and also all the other people that experience problems with CPU detection) please post type of CPU you have ? Thanks.alpha123 wrote: I'd agree with you. It detected 2 threads on my quad, fortunately I saw that before I started testing it on playchess
Guys, maybe a 1.7.1 with proper thread detection and null move bug fix is needed.
Peter
The null move bug will require regression test anyway so although the fix could be quick the testing will take some time and we would wait at least a week or two to collect all the bug reports before to release a mainteinance version.
Miguel
Am I hearing some criticism here? This too simple DannDann Corbit wrote:it's a new and very interesting idea (I never saw it before) that will prune more and more aggressively as a function of existing search depth.Aaron Becker wrote:Here is the new nullmove depth reduction calculation:nepossiver wrote:Was Dann Corbit smooth scaling introduced into the source? I remember the few tests done showed either no difference or a 20-30 increase in ELO (which is what I found in 1+1 games).
I don't know how close this is to Dann's formulation.Code: Select all
// Null move dynamic reduction based on depth int R = 3 + (depth >= 5 * OnePly ? depth / 8 : 0); // Null move dynamic reduction based on value if (refinedValue - beta > PawnValueMidgame) R++;
For instance, if you are at a depth of 32, R would be 3+4=7.
Did you ever see a 7 ply reduction before? It's the most unheard of thing I have ever heard of!
? You want to do a reduction of seven plies only for very promising moves and the rest should be reduced by cos²ln√π(α + Δ(approximateEval - β) + depth) + log(number of days between Easter 2010 and the start of the new Mayan calendar)?