Stockfish

on Fri Mar 14, 2014 7:01 pm by · on Fri Mar 14, 2014 7:01 pm by

@Vladimir wrote:Stockfish 08-03-2014 64bit Gauntlet
http://www.chess2u.com/t7142p75-stockfish-under-microscope#52461

Code:
EnginesWar (13 March 2014) Program Elo + - Games Score Av.Op. Draws 1 Houdini 4 Pro x64 : 3316 9 9 3100 60.1 % 3245 42.0 % 2 Stockfish 12-01-2014 64bit : 3299 13 13 1500 58.6 % 3238 49.1 % 3 Stockfish 08-03-2014 64bit : 3286 12 12 1500 56.9 % 3238 49.7 %

http://www.chess2u.com/t6886p30-engineswar-rating-list#52462

Thank you very much dear Vladimir. thumb up

on Sat Mar 15, 2014 3:38 pm by · on Sat Mar 15, 2014 3:38 pm by

Stockfish dev - Mar 14 2014 : http://abrok.eu/stockfish/
Stockfish dev Syzygy : http://abrok.eu/stockfish_syzygy/

Depth dependant aspiration window delta

Split delta value in aspiration window so that when
search depth is less than 24 a smaller delta value
is used. The idea is that the search is likely to
be more accurate at lower depths and so we can exclude
more possibilities, 25% to be exact.

Passed STC
LLR: 2.96 (-2.94, 2.94) [-1.50, 4.50]
Total: 20430 W: 3775 L: 3618 D: 13037

And LTC
LLR: 2.96 (-2.94, 2.94) [0.00, 6.00]
Total: 5032 W: 839 L: 715 D: 3478

Bench: 7451319

on Mon Mar 17, 2014 10:38 pm by · on Mon Mar 17, 2014 10:38 pm by

@Hossam wrote:Stockfish dev - Mar 14 2014 : http://abrok.eu/stockfish/
Stockfish dev Syzygy : http://abrok.eu/stockfish_syzygy/

Depth dependant aspiration window delta

Split delta value in aspiration window so that when
search depth is less than 24 a smaller delta value
is used. The idea is that the search is likely to
be more accurate at lower depths and so we can exclude
more possibilities, 25% to be exact.

Passed STC
LLR: 2.96 (-2.94, 2.94) [-1.50, 4.50]
Total: 20430 W: 3775 L: 3618 D: 13037

And LTC
LLR: 2.96 (-2.94, 2.94) [0.00, 6.00]
Total: 5032 W: 839 L: 715 D: 3478

Bench: 7451319

I like what Mysseno is doing here.

It seems to me that they have developed their own version of "smart fail-high" for the Stockfish DEV engine! I would speculate that this will add speed to the engine's depth-oriented search algorithm.

on Sun Mar 23, 2014 2:08 pm by · on Sun Mar 23, 2014 2:08 pm by

Stockfish-js
http://www.chess2u.com/t8289-stockfish-js

SPSA Tuner for Stockfish
http://www.chess2u.com/t8290-spsa-tuner-for-stockfish

on Sun Mar 23, 2014 7:37 pm by · on Sun Mar 23, 2014 7:37 pm by

Stockfish dev - Mar 23 2014 : http://abrok.eu/stockfish/
Stockfish dev Syzygy - Mar 23 2014 : http://abrok.eu/stockfish_syzygy/

Refresh TT entries generation automatically on probe

And other assorted simplifications, tested with SPRT[-3, 1]

Passed both short TC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 18814 W: 3600 L: 3475 D: 11739

And long TC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 20731 W: 3217 L: 3096 D: 14418

on Wed Mar 26, 2014 11:45 pm by · on Wed Mar 26, 2014 11:45 pm by

Stockfish dev - Mar 26 2014 : http://abrok.eu/stockfish/

- Fix a bug in pawns eval tracing

on Sat Mar 29, 2014 6:35 am by · on Sat Mar 29, 2014 6:35 am by

Stockfish_ROCKWOOD v. ULTIMA 170214

Link: https://www.dropbox.com/sh/4rubami2nvldcoo/7qhV5SEWxd

on Sun Mar 30, 2014 12:23 am by · on Sun Mar 30, 2014 12:23 am by

Thank you guys! thumb up

on Sun Mar 30, 2014 12:24 am by · on Sun Mar 30, 2014 12:24 am by

@waelsae wrote:Thank you guys!

Welcome

on Sun Mar 30, 2014 12:25 am by · on Sun Mar 30, 2014 12:25 am by

I really like the "Stockwood" Variants!

on Mon Mar 31, 2014 9:13 am by · on Mon Mar 31, 2014 9:13 am by

Hello Folks,
Here's the latest compile from Marco Costalba's StockFish Dev (Fix a compile error with Intel C++) merged with Syzygy & Large Page support. It's compiled using the newest GCC++ 4.8.2, cleaned and optimized for faster engine access (<500kb). Tested with the original dev and RDM versions with excellent results. Enjoy your testing.

http://www.mediafire.com/download/u2ye12cls4g7zuc/stockfish_300314MCx64.7z

Last edited by Ferg2013 on Thu Apr 03, 2014 8:41 am; edited 1 time in total

on Mon Mar 31, 2014 10:22 am by · on Mon Mar 31, 2014 10:22 am by

what s a diferent with http://abrok.eu/stockfish/ ? question

on Mon Mar 31, 2014 10:49 am by · on Mon Mar 31, 2014 10:49 am by

It's the same.

on Mon Mar 31, 2014 11:46 am by · on Mon Mar 31, 2014 11:46 am by

thank's

on Mon Mar 31, 2014 7:40 pm by · on Mon Mar 31, 2014 7:40 pm by

@Ferg2013 wrote:Hello Folks,
Here's the latest compile from Marco Costalba's StockFish Dev (Fix a compile error with Intel C++) merged with Syzygy & Large Page support. It's compiled using the newest GCC++ 4.8.2, cleaned and optimized for faster engine access (<500kb). Tested with the original dev and RDM versions with execellent results. Enjoy your testing.

http://www.mediafire.com/download/u2ye12cls4g7zuc/stockfish_300314MCx64.7z

Thanks you dude.. clap

on Tue Apr 01, 2014 7:33 am by · on Tue Apr 01, 2014 7:33 am by

VocalTechnique wrote:It's the same.

The binary is the same but it has the Syzygy support as well as the Large Page. It is optimized that speed-up engine access. It's strength is not fully verified as I only used a slower cpu during my test. I really appreciate it if anyone has a faster cpu & would like to do further testing.

on Tue Apr 01, 2014 9:00 am by · on Tue Apr 01, 2014 9:00 am by

Hey guys Anyone knows with expirence or searches what is the STRONGEST ENGINE OF STOCKFISH, for long or short games, I`m really confused...

everyday show new engine and a new version... i don`t know more what engine i can use... Tnz for all!!

on Tue Apr 01, 2014 12:30 pm by · on Tue Apr 01, 2014 12:30 pm by

@yuri zarpellon wrote:Hey guys Anyone knows with expirence or searches what is the STRONGEST ENGINE OF STOCKFISH, for long or short games, I`m really confused... everyday show new engine and a new version... i don`t know more what engine i can use... Tnz for all!!

not only you yuri... me to.. VERRY confused...
i used Acer Intel Core i5 430M (2,26GHz) 2 GB memory
with fritz 12 old version no SYSZ
maybe someone can help me
what engine should i use and seting parameter..??
if u dont mind.. plesae give me sollution.. PM me..
I know 2 the best people in this forum are VocalTech and Vladimir... smile

please help me... (sorry my english is not gooed)
thank's before

on Tue Apr 01, 2014 7:05 pm by · on Tue Apr 01, 2014 7:05 pm by

Thanks Ferg2013. thumb up

http://www.chess2u.com/t7485p300-stockfish-tests

on Tue Apr 01, 2014 7:20 pm by · on Tue Apr 01, 2014 7:20 pm by

@Dio wrote:
@yuri zarpellon wrote:Hey guys Anyone knows with expirence or searches what is the STRONGEST ENGINE OF STOCKFISH, for long or short games, I`m really confused...      everyday show new engine and a new version... i don`t know more what engine i can use... Tnz for all!!

not only you yuri... me to.. VERRY confused...
i used Acer Intel Core i5 430M (2,26GHz) 2 GB memory
with fritz 12 old version no SYSZ
maybe someone can help me
what engine should i use and seting parameter..??
if u dont mind.. plesae give me sollution.. PM me..
I know 2 the best people in this forum are VocalTech and Vladimir...
please help me... (sorry my english is not gooed)
thank's before

The latest compile regarding Stockfish DEV at this website:
- http://abrok.eu/stockfish/
- http://abrok.eu/stockfish_syzygy/
Is the strongest.

Unless Marco Costalba, one of the authors, announces a regression, then use last complile before the regression.

Follow the compiles by the dates of compiles.

310314 means March 31, 2014.

Keep tabs on the real-time Stockfish testing, it is a very good indicator of DEV strength.

on Tue Apr 01, 2014 7:35 pm by · on Tue Apr 01, 2014 7:35 pm by

@Ferg2013 wrote:
VocalTechnique wrote:It's the same.

The binary is the same but it has the Syzygy support as well as the Large Page. It is optimized that speed-up engine access. It's strength is not fully verified as I only used a slower cpu during my test. I really appreciate it if anyone has a faster cpu & would like to do further testing.

Robert De Man's compile also supports syzygy and lp. thumb up

on Tue Apr 01, 2014 8:32 pm by · on Tue Apr 01, 2014 8:32 pm by

VocalTechnique wrote:
@Ferg2013 wrote:
VocalTechnique wrote:It's the same.

The binary is the same but it has the Syzygy support as well as the Large Page. It is optimized that speed-up engine access. It's strength is not fully verified as I only used a slower cpu during my test. I really appreciate it if anyone has a faster cpu & would like to do further testing.

Robert De Man's compile also supports syzygy and lp.

I wasn't able to find LP on RDM parameters, pls direct me where to find the LP configuration. And, a "Make Clean Rule" was used to have a smaller file (<500kb).
But all in all, it's comparable to the orig dev & RDM compiles if not better. Thanks

on Tue Apr 01, 2014 9:41 pm by · on Tue Apr 01, 2014 9:41 pm by

RDM - is syzygy. His compiles naturally support syzygy TB - BUT THEY DO NOT EMPLOY/SUPPORT LARGE (Windows)/HUGE (Linux) PAGES. The most popular compiles that support LP are IPMan & Rockwood (Bryce Higgs). Rockwood also support AVX.

on Tue Apr 01, 2014 11:23 pm by · on Tue Apr 01, 2014 11:23 pm by

@Garricar13 wrote:RDM - is syzygy. His compiles naturally support syzygy TB - BUT THEY DO NOT EMPLOY/SUPPORT LARGE (Windows)/HUGE (Linux) PAGES. The most popular compiles that support LP are IPMan & Rockwood (Bryce Higgs). Rockwood also support AVX.

Yet, they are not as strong as the DEV versions (both Robert De Man and Marco Costalba)

on Tue Apr 01, 2014 11:43 pm by · on Tue Apr 01, 2014 11:43 pm by

I agree, DEV versions are more stronger than RW/IP versions.. thumb up