1200 games for DOCH 1.2 added ...

[IWB]

... to the IPON-RATING-LIST: http://www.inwoba.de

Bye
Ingo

[beachknight]

Hi,

I see you are testing spark vc x64 version. Just wanted to note that
spark gcc x64 version is performing much more better in my tests.

Any other comparisons between the two?

Best,

[IWB]

Hi Harun,

I dont know, maybe the programmer can say more but I tested GCC and VC version and (as other engines are faster with VC too) the VC versions has the higher nodes per second in W7x64 and XPx64. So I tested the VC Version.

Bye
Ingo

[beachknight]

Hi Harun,

I dont know, maybe the programmer can say more but I tested GCC and VC version and (as other engines are faster with VC too) the VC versions has the higher nodes per second in W7x64 and XPx64. So I tested the VC Version.

Bye
Ingo

I also employ both, XP x64 and W7 x64. In both OSs, gcc version
performs around 100 points better than the vc version. Well, more
games needed but I think you might want to test the gcc version
with few hundreds games.

Best,

[Allard Siemelink]

Hi Harun,

I dont know, maybe the programmer can say more but I tested GCC and VC version and (as other engines are faster with VC too) the VC versions has the higher nodes per second in W7x64 and XPx64. So I tested the VC Version.

Bye
Ingo

correct, you should use the version that has the highest nps on your machine.

[IWB]

Hi.

I also employ both, XP x64 and W7 x64. In both OSs, gcc version
performs around 100 points better than the vc version.

With all respect, but a different compile on the same code base is highly unlikly to produce 100 Elo difference (and this would bring Sparc on Naum 4 level in my ranking)
If you have a 10% speed difference it might be worth 8 maybe 10 Elo, which is more or less impossible to prove.

But you are right, more games needed.

Bye
Ingo

[beachknight]

Hi.

I also employ both, XP x64 and W7 x64. In both OSs, gcc version
performs around 100 points better than the vc version.

With all respect, but a different compile on the same code base is highly unlikly to produce 100 Elo difference (and this would bring Sparc on Naum 4 level in my ranking)
If you have a 10% speed difference it might be worth 8 maybe 10 Elo, which is more or less impossible to prove.

But you are right, more games needed.

Bye
Ingo

My latest rating list:

http://tech.groups.yahoo.com/group/Ches ... ssage/3608

Very small sample, I agree with you.

Best,