Marvin 3.0.0

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

Moderators: hgm, Rebel, chrisw

Martin
Posts: 51
Joined: Sun Jul 10, 2016 9:12 pm
Location: Lund, Sweden
Full name: Martin Danielsson

Marvin 3.0.0

Post by Martin »

Hi,

I have released Marvin 3.0.0. The main new feature is that Marvin now has multi-thread support using lazy smp. There are also some smaller evaluation and build improvements.

Approximate rating improvement:
1cpu vs 2.2.0: +35 elo
2cpu vs 1cpu: +35 elo
4cpu vs 1cpu: +75 elo

Release: https://github.com/bmdanielsson/marvin- ... tag/v3.0.0
Source code: https://github.com/bmdanielsson/marvin-chess
Top
User avatar
CMCanavessi
Posts: 1142
Joined: Thu Dec 28, 2017 4:06 pm
Location: Argentina

Re: Marvin 3.0.0

Post by CMCanavessi »

Thanx a lot! Another nice update for Season 3 :D
Top
User avatar
Graham Banks
Posts: 41412
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: Marvin 3.0.0

Post by Graham Banks »

Good work Martin. Looking forward to testing it. :)
gbanksnz at gmail.com
Top
User avatar
Mithu
Posts: 213
Joined: Thu Jul 15, 2010 5:59 am

Re: Marvin 3.0.0

Post by Mithu »

Any ELO Rating estimate? How about style? Please let me know if there is any such info available (and where)...
Top
User avatar
Graham Banks
Posts: 41412
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: Marvin 3.0.0

Post by Graham Banks »

Mithu wrote:Any ELO Rating estimate? How about style? Please let me know if there is any such info available (and where)...

Code: Select all

CCRL 40/40 Rating List - Custom engine selection
792542 games played by 2037 programs, run by 21 testers
Ponder off, General books (up to 12 moves), 3-4-5 piece EGTB
Time control: Equivalent to 40 moves in 40 minutes on Athlon 64 X2 4600+ (2.4 GHz), about 15 minutes on a modern Intel CPU.
Computed on February 10, 2018 with Bayeselo based on 792'542 games
Tested by CCRL team, 2005-2018, http://computerchess.org.uk/ccrl/4040/

Rank               Engine                Elo   +    -   Score  AvOp  Games
1 Marvin 2.2.0 64-bit                2721  +25  -25  50.0%   -3.7   524
  Marvin 2.1.0 64-bit                2582  +27  -28  46.5%  +26.6   448
  Marvin 2.0.0 64-bit                2542  +27  -26  59.6%  -70.2   498
  Marvin 1.3.0                       1973  +24  -24  45.0%  +37.1   636
gbanksnz at gmail.com
Top
User avatar
Mithu
Posts: 213
Joined: Thu Jul 15, 2010 5:59 am

Re: Marvin 3.0.0

Post by Mithu »

Graham Banks wrote:
Mithu wrote:Any ELO Rating estimate? How about style? Please let me know if there is any such info available (and where)...

Code: Select all

CCRL 40/40 Rating List - Custom engine selection
792542 games played by 2037 programs, run by 21 testers
Ponder off, General books (up to 12 moves), 3-4-5 piece EGTB
Time control: Equivalent to 40 moves in 40 minutes on Athlon 64 X2 4600+ (2.4 GHz), about 15 minutes on a modern Intel CPU.
Computed on February 10, 2018 with Bayeselo based on 792'542 games
Tested by CCRL team, 2005-2018, http://computerchess.org.uk/ccrl/4040/

Rank               Engine                Elo   +    -   Score  AvOp  Games
1 Marvin 2.2.0 64-bit                2721  +25  -25  50.0%   -3.7   524
  Marvin 2.1.0 64-bit                2582  +27  -28  46.5%  +26.6   448
  Marvin 2.0.0 64-bit                2542  +27  -26  59.6%  -70.2   498
  Marvin 1.3.0                       1973  +24  -24  45.0%  +37.1   636
Thanks for the info Graham! I appreciate your time and effort. Any idea about its style?
Top
User avatar
Graham Banks
Posts: 41412
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: Marvin 3.0.0

Post by Graham Banks »

Mithu wrote:
Graham Banks wrote:
Mithu wrote:Any ELO Rating estimate? How about style? Please let me know if there is any such info available (and where)...

Code: Select all

CCRL 40/40 Rating List - Custom engine selection
792542 games played by 2037 programs, run by 21 testers
Ponder off, General books (up to 12 moves), 3-4-5 piece EGTB
Time control: Equivalent to 40 moves in 40 minutes on Athlon 64 X2 4600+ (2.4 GHz), about 15 minutes on a modern Intel CPU.
Computed on February 10, 2018 with Bayeselo based on 792'542 games
Tested by CCRL team, 2005-2018, http://computerchess.org.uk/ccrl/4040/

Rank               Engine                Elo   +    -   Score  AvOp  Games
1 Marvin 2.2.0 64-bit                2721  +25  -25  50.0%   -3.7   524
  Marvin 2.1.0 64-bit                2582  +27  -28  46.5%  +26.6   448
  Marvin 2.0.0 64-bit                2542  +27  -26  59.6%  -70.2   498
  Marvin 1.3.0                       1973  +24  -24  45.0%  +37.1   636
Thanks for the info Graham! I appreciate your time and effort. Any idea about its style?
I can't recall offhand. Sorry. :(
gbanksnz at gmail.com
Top
tpoppins
Posts: 919
Joined: Tue Nov 24, 2015 9:11 pm
Location: upstate

Re: Marvin 3.0.0

Post by tpoppins »

You can download hundreds of 40/40 games of various Marvin versions at CCRL, form your own conclusions about its style and share them here.
Top
LocutusOfPenguin
Posts: 34
Joined: Thu Sep 28, 2017 6:52 pm
Location: Karlsruhe, Germany
Full name: Jürgen Précour

Re: Marvin 3.0.0

Post by LocutusOfPenguin »

Hi,

- in yr readme.md y have a typo at building > comaptible
- if i type "go movetime 50" or smaller i get a "a1a1" from 51 onwards, it calcs and print e2e4

Jürgen
Create a dedicated chess computer based on tiny ARM computers with the DGT e-board on picochess.com
Top
Martin
Posts: 51
Joined: Sun Jul 10, 2016 9:12 pm
Location: Lund, Sweden
Full name: Martin Danielsson

Re: Marvin 3.0.0

Post by Martin »

Thank you for reporting these issues. I'll make sure to fix them for the next release.
Top

Return to “Computer Chess Club: General Topics”