Fruit 05/11/03 tourney

[IWB]

Hello all

Little tourney to have a better stength estimation for Fruit 05/11/03

Fruit 05/11/03 - Deep Shredder 10 x64 1T 53.5- 46.5 53.50%
Fruit 05/11/03 - Rybka 1.0 Beta 32-bit 49.0- 51.0 49.00%
Fruit 05/11/03 - Toga II 1.2.1a 46.5- 53.5 46.50%
Fruit 05/11/03 - Spike 1.2 Turin 61.0- 39.0 61.00%
Fruit 05/11/03 - Loop 13.5 53.0- 47.0 53.00%
Fruit 05/11/03 - HIARCS 11.1 NoLearn 52.0- 48.0 52.00%
315.0-285.0 52.50%

600 out of 600 games played

50 Nunn2 position canging colors
1 Thread
PONDER ON (!)
256 MB Hash
2300 MHz XEON 5320 (Core 2 core)
6min + 3 sec (approx. 16 min per game)

The only thing that does not fit is the score vs. its brother Toga. Overall a very nice Engine, I am impressed!

Bye
Ingo

[George Tsavdaris]

Hello all

Little tourney to have a better stength estimation for Fruit 05/11/03

Fruit 05/11/03 - Deep Shredder 10 x64 1T 53.5- 46.5 53.50%
Fruit 05/11/03 - Rybka 1.0 Beta 32-bit 49.0- 51.0 49.00%
Fruit 05/11/03 - Toga II 1.2.1a 46.5- 53.5 46.50%
Fruit 05/11/03 - Spike 1.2 Turin 61.0- 39.0 61.00%
Fruit 05/11/03 - Loop 13.5 53.0- 47.0 53.00%
Fruit 05/11/03 - HIARCS 11.1 NoLearn 52.0- 48.0 52.00%
315.0-285.0 52.50%

600 out of 600 games played

50 Nunn2 position canging colors
1 Thread
PONDER ON (!)
256 MB Hash
2300 MHz XEON 5320 (Core 2 core)
6min + 3 sec (approx. 16 min per game)

The only thing that does not fit is the score vs. its brother Toga. Overall a very nice Engine, I am impressed!

Wow, beating Hiarcs 11.1, Shredder 10, Loop 13.5 and Spike 1.2 is something amazing for a free engine. Perhaps they've made a mistake for releasing it for free....

Can you send me the games or provide a place to download them?
gtsavdar@gmail.com is my email.

[Tony Thomas]

I think that Fruit 2.3 might even be stronger. I only have around 60 games so far, but it is rated about 60 points higher than the free Rybka so far. Of course I dont have enough games, but I can say for sure that it is no push over.

[George Tsavdaris]

I think that Fruit 2.3 might even be stronger. I only have around 60 games so far, but it is rated about 60 points higher than the free Rybka so far. Of course I dont have enough games, but I can say for sure that it is no push over.

Any idea what Fruit is stronger from all these in the new file?
I have:
Fruit 2.3
Fruit23-EM64T (what is this?)
Fruit-070327
Fruit-051103
Fruit-061115a
Fruit-061115b

Any description about all these or a log for the changes.....?
Which seems to be stronger from all these?

I know that the last question is difficult to answer for now, but i still hope for an answer....

Also thanks to Ryan for doing all these. It seems from these first matches that Fruit is again on top places....

[Tony Thomas]

I think that Fruit 2.3 might even be stronger. I only have around 60 games so far, but it is rated about 60 points higher than the free Rybka so far. Of course I dont have enough games, but I can say for sure that it is no push over.

Any idea what Fruit is stronger from all these in the new file?
I have:
Fruit 2.3
Fruit23-EM64T (what is this?)
Fruit-070327
Fruit-051103
Fruit-061115a
Fruit-061115b

Any description about all these or a log for the changes.....?
Which seems to be stronger from all these?

I know that the last question is difficult to answer for now, but i still hope for an answer....

Also thanks to Ryan for doing all these. It seems from these first matches that Fruit is again on top places....

Fruit 2.3 is the latest version, I always use the latest version before I play around with the betas. The em64t is a 64bit version optimized for intel. The last two versions are the ones that Thorsten always brag about (I think).

[IWB]

Hello

I think that Fruit 2.3 might even be stronger. I only have around 60 games so far, but it is rated about 60 points higher than the free Rybka so far. Of course I dont have enough games, but I can say for sure that it is no push over.

Any idea what Fruit is stronger from all these in the new file?
I have:
Fruit 2.3
Fruit23-EM64T (what is this?)
Fruit-070327
Fruit-051103
Fruit-061115a
Fruit-061115b

Any description about all these or a log for the changes.....?
Which seems to be stronger from all these?

I know that the last question is difficult to answer for now, but i still hope for an answer....

Also thanks to Ryan for doing all these. It seems from these first matches that Fruit is again on top places....

Fruit 2.3 is the latest version, I always use the latest version before I play around with the betas. The em64t is a 64bit version optimized for intel. The last two versions are the ones that Thorsten always brag about (I think).

If I understand it correctly Fruit 2.3 is not developed by Fabien and it is "just" from the last public Fruit sourcecode.

Fruit 05/11/03 is the "Last Fabien Version" ( I call it internaly LFV)
All others are developed by someone else and the best one is not determinded yet!

Unfortunately Fabien allowed 2.3 to be named "Fruit" as well and now we have a lot of confusion.

EM64T is just a 64bit OS compile. Sometimes they are a bit faster on 64bit OSs then the normal 32 bit versions. The sourcecode is the same.

In General I agree with you about the latest Version, in this case it is different and somehow special.

BYe
Ingo

[Ryan Benitez]

Fruit 2.3 was not built from the last open source version. I did type the code from Fruit the version of Fruit that I was sent by Fabien, was in communication with Fabien, and he did review code from time to time.

Thanks,
Ryan

[IWB]

Fruit 2.3 was not built from the last open source version. I did type the code from Fruit the version of Fruit that I was sent by Fabien, was in communication with Fabien, and he did review code from time to time.

Thanks,
Ryan

Hello

That is interesting to hear. Have you done some testing if the 2.3 is "better" then the 05/11/03?

Thx 4 your work
Ingo

[Eelco de Groot]

Fruit 2.3 was not built from the last open source version. I did type the code from Fruit the version of Fruit that I was sent by Fabien, was in communication with Fabien, and he did review code from time to time.

Thanks,
Ryan

Hello Ryan,

Thanks a lot for all the different Fruit versions! I hope it does not mean that you have no more time for chessprogramming though. I was also curious, since you already seemed quite far with a parallel version of Fruit for the tournament in Leiden Ryan, did you stop working on it because Fabien was doing a rewrite of his version, or simply it never got done the way you wanted it to work?

Also, the most obvious thing to do it seems judging from the stories about Osipov's Strelka mainly, apart maybe introducing some bitboard presentation of things, would appear to be a better evaluation of material imbalances. I have not really studied any games of Fruit in that respect, how well Fruit is doing that compared to older versions, I was just wondering if you think that would help Fruit also. Maybe such knowledge only becomes feasible with a bitboard presentation, I have not really thought about it, if it is very difficult to program without bitboards.

Thanks again Ryan and to Fabien too, I hope the Fruit project can continue!

Eelco

[Bryan Hofmann]

Fruit23-EM64T (what is this?)

This is a 64 bit compile (optimized for Intel Chips) created with the Orcas Beta compiler from Micosoft. You fill find it a shade slower then the 32 bit compile.


Bryan