1650 games are finished. I put a preliminary result on http://www.inwoba.de
PS: Sorry for not transmitting, but that is difficult with CB native engines.
Based on your list, Deep Fritz 13 1CPU is only 3-4 Elo points stronger than Fritz 13. This means there are not improvement at all over Fritz 13.
Edit: Ok, still I see there are missing games. We'll see what happens.
I though that Deep Fritz 13 in one core, was a slighty different/improved version of Fritz 13.
> I though that Deep Fritz 13 in one core, was a slighty different/improved version of Fritz 13.
usually it is.
>I though that Deep Fritz 13 in one core, was a slighty different/improved version of Fritz 13.
That was the case until F/DF12 but apperantly it is different with F/DF13. On the other had it is now what all others "Deep" engines already are ...
Additionaly my set of games is missing all the "weaker" engines in my usuall test setup right now, while Fritz 13 played against them. We have to wait until the complete run is finished.
Deep Fritz 13 increased its rating significantly over the 1650 games result. It show quite nice that it is not good to publish results of a small sub group!
The IPON-RRRL will be updated later today.
All single core computers are dual nowdays so we can never get exactly the same conditions for both programs.
4x Intel Core i7 CPU X980 (24x4600 MHz) 64GB RAM :
Time 120'+60'+60' . Ponder off . GUI Deep Fritz 13 . Hash 16GB. Nalimov 7 men Lomonosov TB's. Book: HS-7moves.ctg:
Deep Fritz 13 - Houdini 2.0c x64 24 threads :
59/100 - 41/100 . 59%-41% . ( +25 - 7 = 68 )
Deep Fritz 13 is the strongest engine of all on my hardware.
It is using 4 cpu and long time. Someone should do the similar test to see if it happens again.
> GUI Deep Fritz 13 . Hash 16GB. Nalimov 7 men Lomonosov TB's.
Fritz can't (and perhaps never will) access Lomonosov tablebases. You are easily fooled, siah.
Of course multiple might have been used. Or it might be a joke.
Of course being newer it reaches higher depth faster but it's still inferior tactically for some reason.
Is there evidence of a rehash?
28 Rybka 2.3.2a 64-bit – Fritz 11 68.3 2268
Fritz 13 is still the same engine but improved of course.
Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill