A privat RobboLito 0.21Q test under CEGT-Blitz conditions

[ThatsIt]

Hi to all !

Version = RobboLito 0.21Q x64 SSE4.2 NLP

Just have a look at:
http://cegt.siteboard.eu/viewtopic.php? ... p=876#p876

Best wishes,
G.S.

[geots]

Hi to all !

Version = RobboLito 0.21Q x64 SSE4.2 NLP

Just have a look at:
http://cegt.siteboard.eu/viewtopic.php? ... p=876#p876

Best wishes,
G.S.

Thank you Gerhard, for sharing a very interesting match/matches with us. This is the 1st actual results from 0.21Q I have seen from a pro tester like yourself. If some other good ones have posted results- my apologies. I guess I just missed them. I'm so wrapped up with RRs and a new system- I don't right now have time to do anything. A WORD OF CAUTION: Take NLP off the end of the name by using the other executable- the x64 SSE4.2. It will play exactly the same way when you don't even access large pages. Tho maybe you don't need to. Only if you try a RR in a chessbase gui. Then with "NLP" there, it is a coin flip as to whether Fritz 11 or 13 will let it play in a RR with that long a name. It is one of those deals that cannot hurt- but very likely may help.


Best,

george

[ThatsIt]

Hi George !

All the matches were done under the Shredder-Classic-GUI.
There is no need to use a CB-GUI at all

Best wishes,
G.S.

[Dragan]

It doesn't look like a big improvement over 0.09 version, but I may be wrong.
It would be interesting repeating the test with 0.09
Thx, Dragan

[ThatsIt]

Just 200 games are played, let's wait and see...

Best wishes,
G.S.

[ThatsIt]

Hi to all !

Version = RobboLito 0.21Q x64 SSE4.2 NLP

Just have a look at:
http://cegt.siteboard.eu/viewtopic.php? ... p=876#p876

Best wishes,
G.S.

400 games played ...

Best wishes,
G.S.

[ThatsIt]

Hi to all !

Version = RobboLito 0.21Q x64 SSE4.2 NLP

Just have a look at:
http://cegt.siteboard.eu/viewtopic.php? ... p=876#p876

Best wishes,
G.S.

600 games played ...

Best wishes,
G.S.

[ThatsIt]

Hi to all !

700 games played:
http://cegt.siteboard.eu/viewtopic.php? ... p=876#p876

RobboLito is surely not a "TOP-3" engine as (too) often told.
Wasted time, test finished.

Best wishes,
G.S.

[Houdini]

Similar test results here: http://www.chess2u.com/t6341-robbolito- ... ames#37681 .

[Tomcass]

Similar test results here: http://www.chess2u.com/t6341-robbolito- ... ames#37681 .

My results are different, Robert. In my test, with 6 real cores, Robbolito 0.21Qi is clearly better than Robbolito 0.20Qi. Only 240 games, though. So I have to be cautious.

Testing Robbolito 0.20Q = 120 games
I7 980 3.33 Ghz.
6 real cores
GUI: Fritz 12
Book: Perfect 2012
Time control: 2 minutes +2 sec/ game.
Ponder: Off
No tablebases

201209Robbolito 0.20Q

RobboLito 0.20Q x64 SSE4.2_6-Critter 1.6 64-bitx6_nob 16.0 - 24.0 +3/=26/-11 40.00%
RobboLito 0.20Q x64 SSE4.2_6-Deep Rybka 4.1 SSE42 x64 (x6) 21.0 - 19.0 +8/=26/-6 52.50%
RobboLito 0.20Q x64 SSE4.2_6 - Houdini 2.0c Pro x64(x6) 11.5 - 28.5 +5/=13/-22 28.75%

Average score after 120 games: 40.42%

Games: http://www.mediafire.com/?5jsaycwu8sa9558


Testing Robbolito 0.21Q = 120 games
I7 980 3.33 Ghz.
6 real cores
GUI: Fritz 12
Book: Perfect 2012
Time control: 2 minutes +2 sec/ game.
Ponder: Off
No tablebases

201209Robbolito 0.21Q
RobboLito 0.21Q x64 SSE4.2- Critter 1.6a 64-bit_x6 17.0 - 23.0 +5/=24/-11 42.50%
RobboLito 0.21Q x64 SSE4.2- Deep Rybka 4.1 SSE42 x64 (x6) 18.0 - 22.0 +7/=22/-11 45.00%
RobboLito 0.21Q x64 SSE4.2- Houdini 2.0c Pro x64(x6) 17.0 - 23.0 +8/=18/-14 42.50%

Average score: 43.33%

Robbolito 0.20Q scored 40.42% in the same test. A good improvement!.

Games: http://www.mediafire.com/?b2air85743mg4vc

Regards,

Tom

... I will test again both compiles in the same conditions, only to minimize statistical randomness.