Homayoun wrote: ↑Fri Apr 05, 2024 11:26 pm
Dear Sedat
Thanks also for your rating list. Yes it seems that author of Plentychess e.g exactly knows what he wants from chess programming. Just after 3 test prereleases now we have the main release (No1) among top 10 engines. According to Stefan pohl tests now Berserk 13 is stronger than Komodo dragon, but seems that Berserk couldn’t prove itself in your tests, yet. I am also waiting for CCRL results too.
Greetings
Not at all..
You may know,
Any engine performance depends on our used cond.
So here, I think that the Elo differences are related
Mainly to hardware, e.g SPCC is using HT is enabled..
Where I use as HT disabled, I mean my tested engines
Are used to play with real physical cores...e.g, There is
BIG speed diff. between Logical threads vs Physical cores
Other difference: SCCT's used openings (Balsa) are much
Stronger than SPCC (UHO), so some engines seems to be
Stronger via stronger openings, where some just opposite..
To be sure, you can see K.Dragon 3.3 vs Stochfish 16.1
In other words, under SCCT: the Elo difference is much less,
Where under SPCC cond. we see too much Elo differences!
Sure there can be other factors...such as we used to play
Not same TC, different hashtable sizes etc. but I say again
The main influence is playing hardware, plus openings!
Greetings )