Testing new Ivans (JR and Peterpan compiles)

[Tomcass]

TEST 1
2010-06IvanT57D-1 2010
120 GAMES
Quad 2,33
Gui: Fritz 12
Book: HS 2.0 Masterbook
Time Control: 4 min + 2 sec.
Ponder: Off
1 2 3 4
1 Houdini 1.02 w32 4_CPU 11.0 - 9.0 10.5 - 9.5 13.0 - 7.0** 34.5/60
2 IvanHoe 57a2JR 9.0 - 11.0 11.0 - 9.010.0 - 10.0 ** 30.0/60
3 Ivanhoe-B57dSSE1x32Peterpan 9.5 - 10.5 9.0 - 11.0 10.5 - 9.5 ** 29.0/60
4 Deep Rybka 4 w32 7.0 - 13.010.0 - 10.09.5 - 10.5 ** 26.5/60


TEST 2

201007IvanPeterpan 2010
120 GAMES
I7 975
Gui: Fritz 12
Book: HS 1.0 Masterbook
Time Control: 4 min + 2 sec.
Ponder: Off



1 2 3 4
1 Houdini DEVEL x64 8_CPU 10.0 - 10.09.0 - 11.011.5 - 8.5** 30.5/60
2 IvanHoe 57a2 JR 10.0 - 10.010.0 - 10.010.0 - 10.0 ** 30.0/60 0.00
3 Ivanhoe-B57dSSE4Peterpan 11.0 - 9.010.0 - 10.09.0 - 11.0 ** 30.0/60 0.00
4 Deep Rybka 4 x64 8.5 - 11.510.0 - 10.011.0 - 9.0 ** 29.5/60

GLOBAL RESULTS

240 GAMES.
1 Houdini 1.02 65,0/120
2 IvanHoe 57a2JR 60.0/120
3 Ivanhoe-B57dPeterpan 59.0/120
4 Deep Rybka 4 56,0/120

Houdini 1.02 again the best at short time controls. The latest IVANS compiled by JR and Peterpan are extremely strong.

Regards,

TOM.

[pogi]

Good day Tom, I am not sure if I should reply to you or if I should just post a new topic. I am new here and wondering where can I get help regarding Ivanhoe/Deep Rybka 4/Fire/etc parameter settings. Is there a tutorial anywhere that explains those items that pop up when you click on Engine Parameters (Chessbase GUI)? I want to learn what those items mean and how they affect the level of play. I'd appreciate it if someone can point me to the right direction to learn it.

Thank you.

[Tomcass]

Hello, Romel!

It is a bit difficult for me to give you a guidance on this topic. Regarding to Rybka settings I suggest you to follow any of the trends talking about Deep Rybka 4. Some new settings have been claimed as stronger than the original ones, but I am not sure about it.

Regarding Ippo engines (Houdini, Ivanhoes, Fire and similar) links are not authorised in this forum because some people -the forum sponsors!- believe that they are Rybka clones. (My opinion is very clear: They are not Rybka clones at all, as have been vastly demonstrated by other voices much more professional than mine). But if you are in this forum you have to respect the internal rules.

You will find some information in the 'forbiden side' of talkchess.com : Engine Origins. For any other assistance please do not hesitate to contact me privately.

Regards from Barcelona.

Tom.

[Tomcass]

TEST 1
120 GAMES
Quad 2,33 All engines using 4 cores
Time control 5 min / 0
Gui: Fritz 12
Book: HS Masterbook 2.0
Ponder: Off

2010-07Ivan55b-1 2010

IvanHoe 9.55b w32 (x4) - Deep Rybka 4 w32 34.0 - 26.0 +15/=38/-7 56.67%
IvanHoe 9.55b w32 (x4) - IvanHoe 57a4 JR 30.5 - 29.5 +7/=47/-6 50.83%


TEST 2
120 GAMES
I7 975 All engines using 8 threads (4 cores + hyperthread)
Time control 5 min / 0
Gui: Fritz 12
Book: HS Masterbook 1.0
Ponder: Off

201007Ivan999955m 2010

IvanHoe 9.55b x64 popcnt (x8) - Deep Rybka 4 x64 30.5 - 29.5 +13/=35/-12 50.83%
IvanHoe 9.55b x64 popcnt (x8) - IvanHoe 57a4 (x8) 32.5 - 27.5 +9/=47/-4 54.17%



GLOBAL
240 GAMES
IvanHoe 9.55b - Deep Rybka 4 64.5 – 55.5
IvanHoe 9.55b - IvanHoe 57a4 JR 63.0 - 57.0

Very promising results for this new superstrong engine!

Tom.

[gerold]

Thanks for your work Tom.
Are you using TB for Rybka 4.
In 5 minute games Houdini looks stronger than R4.
In long TC if Rybka is using TB Rybka 4 appears to
be a little stronger than Houdini. Just to few games
to tell yet.

Best,
Gerold.

[Tomcass]

Thanks for your work Tom.
Are you using TB for Rybka 4.
In 5 minute games Houdini looks stronger than R4.
In long TC if Rybka is using TB Rybka 4 appears to
be a little stronger than Houdini. Just to few games
to tell yet.

Best,
Gerold.

Thanks for your comments, Gerold. My findings totally agree with yours. At tournament time control, Deep Rybka seems to be stronger.

In my tests I do not use Nalimov or Robbo tablebases for any engine.

Regards,

Tom.