Komodo 10.2 released.

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

lkaufman
Posts: 5959
Joined: Sun Jan 10, 2010 6:15 am
Location: Maryland USA

Komodo 10.2 released.

Post by lkaufman »

We have released komodo 10.2 at komodochess.com. We estimate based on our testing that it is about 22 elo stronger than komodo 10.1 and about 30 elo stronger than komodo 10 at three minutes plus one second increment, on one or many threads. This is the largest elo gain we have had between versions since going to the subscription model. There were so many changes that it is very difficult to say which ones should get most of the credit. Evaluation, search, and time management were all changed significantly. Features (UCI options) are the same as in komodo 10.1. As usual, it is free to subscribers, and available at a 20% discount to those who bought komodo 9 or a later version.
Komodo rules!
TShackel
Posts: 313
Joined: Sat Apr 05, 2014 12:09 am
Location: Neenah, WI, United States

Re: Komodo 10.2 released.

Post by TShackel »

Thanks much Larry! Congrats on the huge improvement! Can't wait to start testing it. It played very well in the rapid tournament. Keep up great work.

Sincerely,

Tim Shackel.
Frank Quisinsky
Posts: 6808
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: Komodo 10.2 vs. ASM Fish ... started Monday!

Post by Frank Quisinsky »

Hi Larry,

OK, this will be the last "Fight for Place 1" from my site.

Komodo 10.2 vs. ASM Fish current version.
Both vs. 49 opponents to the same time.

Will be start Monday in the very late evening for the next 11 days!

I must see Komodo 10.2 and ASM before my list ended and before I moving around Dec. 15th, 2016.

Computer chess can be very hard ... but I love it.
Means with the work I have with it.

Good news!

Please tell me ...
Which setting for 40 moves on 10 minutes on 4Ghz i7 hardware with 1 core and 4-pieces you prefer. I will give Komodo 10.2 your settings.

Thanks for the new version and regards to your partner!

Best
Frank

PS:
At the moment Nemorino is still running ... very strong ... nice new engine!
User avatar
MikeB
Posts: 4889
Joined: Thu Mar 09, 2006 6:34 am
Location: Pen Argyl, Pennsylvania

Re: Komodo 10.2 released.

Post by MikeB »

lkaufman wrote:We have released komodo 10.2 at komodochess.com. We estimate based on our testing that it is about 22 elo stronger than komodo 10.1 and about 30 elo stronger than komodo 10 at three minutes plus one second increment, on one or many threads. This is the largest elo gain we have had between versions since going to the subscription model. There were so many changes that it is very difficult to say which ones should get most of the credit. Evaluation, search, and time management were all changed significantly. Features (UCI options) are the same as in komodo 10.1. As usual, it is free to subscribers, and available at a 20% discount to those who bought komodo 9 or a later version.
Very impressive - will add the the latest to my engine stable and start testing immediately!
Modern Times
Posts: 3539
Joined: Thu Jun 07, 2012 11:02 pm

Re: Komodo 10.2 released.

Post by Modern Times »

Not easy to get that sort of Elo improvement at this level, great work !
Frank Quisinsky
Posts: 6808
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: Komodo 10.2 released.

Post by Frank Quisinsky »

We wish to throw no one into the shade, but we demand our own place in the sun.

Nice time computer chess have!

Most important question for myself:
How many win games undo move 60 undo mate ... I am sure new records will be the result from both engines.

So nice to see how aggressive Komodo and Stockfish can be play.
My big compliment to the team of Komodo and to the team of Stockfish.

In reality ...
Komodo and Stockfish have the place in the sun.
And I the sunburn.

Best
Frank
lkaufman
Posts: 5959
Joined: Sun Jan 10, 2010 6:15 am
Location: Maryland USA

Re: Komodo 10.2 vs. ASM Fish ... started Monday!

Post by lkaufman »

Frank Quisinsky wrote:Hi Larry,

OK, this will be the last "Fight for Place 1" from my site.

Komodo 10.2 vs. ASM Fish current version.
Both vs. 49 opponents to the same time.

Will be start Monday in the very late evening for the next 11 days!

I must see Komodo 10.2 and ASM before my list ended and before I moving around Dec. 15th, 2016.

Computer chess can be very hard ... but I love it.
Means with the work I have with it.

Good news!

Please tell me ...
Which setting for 40 moves on 10 minutes on 4Ghz i7 hardware with 1 core and 4-pieces you prefer. I will give Komodo 10.2 your settings.

Thanks for the new version and regards to your partner!

Best
Frank

PS:
At the moment Nemorino is still running ... very strong ... nice new engine!
The defaults are probably fine for your purpose, except of course Hash size which I suppose you will decide. Using a power of two slightly favors Stockfish (we prefer 3x a power of 2) but probably you have a set rule for this. Table memory could be doubled to 256 but it's not very clear or very important. Due to the large number (and therefore range of strength) of opponents we would probably do better with Contempt 15 than with the default 10, but if you want to keep it at 10 to match Stockfish I won't complain. If you ever get any time forfeits you can raise the Overhead ms from 30 to 50, but probably you never get forfeits at the level you use.

Best,
Larry
Komodo rules!
Frank Quisinsky
Posts: 6808
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: Komodo 10.2 vs. ASM Fish ... started Monday!

Post by Frank Quisinsky »

Hi Larry,

great will looking in the night and try it out.
I think why not ... Contempt = 15 ... but I think I should set it for SF in this case too.

Never I have forfeits problems with Komodo. Komodo runs all the time without any technical problems here. Each of the tested Komodo versions. Hash I think with 512Mb is OK, all are playing with 512Mb all the time.

OK Contempt = 15!
You are the chief!

Thanks again!

Best
Frank
lkaufman
Posts: 5959
Joined: Sun Jan 10, 2010 6:15 am
Location: Maryland USA

Re: Komodo 10.2 vs. ASM Fish ... started Monday!

Post by lkaufman »

Frank Quisinsky wrote:Hi Larry,

great will looking in the night and try it out.
I think why not ... Contempt = 15 ... but I think I should set it for SF in this case too.

Never I have forfeits problems with Komodo. Komodo runs all the time without any technical problems here. Each of the tested Komodo versions. Hash I think with 512Mb is OK, all are playing with 512Mb all the time.

OK Contempt = 15!
You are the chief!

Thanks again!

Best

I agree it's fair to set Stockfish to Contempt 15 if we do so, although Contempt has much greater consequences in Komodo than in Stockfish last I looked.
Frank
Komodo rules!
S.Taylor
Posts: 8514
Joined: Thu Mar 09, 2006 3:25 am
Location: Jerusalem Israel

Re: Komodo 10.2 released.

Post by S.Taylor »

lkaufman wrote:We have released komodo 10.2 at komodochess.com. We estimate based on our testing that it is about 22 elo stronger than komodo 10.1 and about 30 elo stronger than komodo 10 at three minutes plus one second increment, on one or many threads. This is the largest elo gain we have had between versions since going to the subscription model. There were so many changes that it is very difficult to say which ones should get most of the credit. Evaluation, search, and time management were all changed significantly. Features (UCI options) are the same as in komodo 10.1. As usual, it is free to subscribers, and available at a 20% discount to those who bought komodo 9 or a later version.
What _I_ want to see is how this new komodo 10.2 does vs the SF that played in stage 3 of TCEC, OR, the version of SF that played in Graham Banks matches between sf and komodo.

It is very frustrating to have two moving targets at the same time.

e.g. let's say komodo beats ASM, how will we know if the older SF version, which Graham used, wouldn't have crushed komodo 10.2 like it did to komodo 10.1?

We should first test a moving target with a standing target before we go on to the second moving target.