Bouquet 1.4 vs Zappa Mexico II- FINAL RESULTS!

[geots]

Bouquet_1.4pop x64 vs Zappa Mexico II x64


Look, right now there are too many questions and no answers. Before I can judge the performance of Bouquet- to be fair, I would have to know more than I do now. Starting in August of 2009, I was out of the loop for about a year and a half. I can't imagine in that time that Cozzie released another 64bit exe for Zappa Mexico II, as it wasn't even new then. And he never cared one way or the other when it was new- that I could tell. So I would have to assume the 64bit exe I have now is the only that ever existed. But I wish someone would help me on that- maybe Sedat. He has Zappa Mexico rated number 21 on his list- tho I admit I did not look at the number of cores it was using there.

Problem is where Bouquet is averaging around depth 19 or 20 , Zappa Mexico is averaging 12 to 13 in ply depth. About 5 to 8 or 9 ply less in each game than Bouquet. More than once he did not even average a ply depth of 12! A crappy 32bit program does better than that. The following is about the norm for Zappa:


{Intel(R) Core(TM) i5-2320 CPU @ 3.00GHz 2993 MHz W=18.3 plies; 2,745kN/s;
Perfect2012b.ctg B=12.9 plies; 1,032kN/s; Perfect2012b.ctg}


or- if you like it worse:

{Intel(R) Core(TM) i5-2320 CPU @ 3.00GHz 2993 MHz W=11.9 plies; 888kN/s;
Perfect2012b.ctg B=18.2 plies; 2,443kN/s; Perfect2012b.ctg}

Please don't ask me which is which. Because I don't really see how a 64bit engine could win at this rate. So I certainly congratulate Velmarin and Bouquet. No doubt he won. Problem is- based on what he seems to have beat- I can't judge the performance. And I know Velmarin understands that. And who knows- it may be the best Zappa can do.

Maybe someone can help me here. Is this the best kns and ply depth Zappa can come up with? But to the results:


Intel i5 w/4TCs
Fritz 13 gui
1CPU/64bit
128MB hash
Bases=NONE
Ponder_Learning=OFF
Perfect 2012b.ctg w/12-move limit
40/3 Repeating (Benched to adapt to 40/4)
Match=50 games

Code: Select all

Bouquet_1.4pop x64     +164    +26/-4/=20   72.00%   36.0/50
Zappa Mexico II x64    -164    +4/-26/=20   28.00%   14.0/50

What can I say except good work to Bouquet!



george

[geots]

Bouquet_1.4pop x64 vs Zappa Mexico II x64


Look, right now there are too many questions and no answers. Before I can judge the performance of Bouquet- to be fair, I would have to know more than I do now. Starting in August of 2009, I was out of the loop for about a year and a half. I can't imagine in that time that Cozzie released another 64bit exe for Zappa Mexico II, as it wasn't even new then. And he never cared one way or the other when it was new- that I could tell. So I would have to assume the 64bit exe I have now is the only that ever existed. But I wish someone would help me on that- maybe Sedat. He has Zappa Mexico rated number 21 on his list- tho I admit I did not look at the number of cores it was using there.

Problem is where Bouquet is averaging around depth 19 or 20 , Zappa Mexico is averaging 12 to 13 in ply depth. About 5 to 8 or 9 ply less in each game than Bouquet. More than once he did not even average a ply depth of 12! A crappy 32bit program does better than that. The following is about the norm for Zappa:


{Intel(R) Core(TM) i5-2320 CPU @ 3.00GHz 2993 MHz W=18.3 plies; 2,745kN/s;
Perfect2012b.ctg B=12.9 plies; 1,032kN/s; Perfect2012b.ctg}


or- if you like it worse:

{Intel(R) Core(TM) i5-2320 CPU @ 3.00GHz 2993 MHz W=11.9 plies; 888kN/s;
Perfect2012b.ctg B=18.2 plies; 2,443kN/s; Perfect2012b.ctg}

Please don't ask me which is which. Because I don't really see how a 64bit engine could win at this rate. So I certainly congratulate Velmarin and Bouquet. No doubt he won. Problem is- based on what he seems to have beat- I can't judge the performance. And I know Velmarin understands that. And who knows- it may be the best Zappa can do.

Maybe someone can help me here. Is this the best kns and ply depth Zappa can come up with? But to the results:


Intel i5 w/4TCs
Fritz 13 gui
1CPU/64bit
128MB hash
Bases=NONE
Ponder_Learning=OFF
Perfect 2012b.ctg w/12-move limit
40/3 Repeating (Benched to adapt to 40/4)
Match=50 games

Code: Select all

Bouquet_1.4pop x64     +164    +26/-4/=20   72.00%   36.0/50
Zappa Mexico II x64    -164    +4/-26/=20   28.00%   14.0/50

What can I say except good work to Bouquet!



george

I checked with 3 people who everyone would admit would know. And all is well with Zappa Mexico II 64bit. It is just old and slow- with none of the new technology. I did not check for myself. But Velmarin is my friend, and I will be damned if I was going to let his victory have a cloud hanging over it.

So a fair victory! And well done! You demolished an engine that was ahead of your 1.2- even tho 1.2 only had 60 or so games. Very good work.


Below are the PGNs still available for download:


Houdini- Komodo 5, 200 games at 40/4 (Link good till Aug.1)

http://dl.dropbox.com/u/76847541/Houdin ... t_PGNs.rar



Houdini- Komodo5, 50 games at 40/40 (Link good till Aug.1)

http://dl.dropbox.com/u/76847541/Houdin ... ile%5D.rar



Bouquet 1.4- Zappa Mexico II, 50 games at 40/4 (Link good till Aug. 6)

http://dl.dropbox.com/u/76847541/Bouque ... %20x64.rar



Enjoy-

george

[Sedat Canbaz]

Thanks for your interesting testing dear George

It seems,Bouquet 1.4 is a very strong engine...

But unfortunately,i could not run Bouquet 1.4 on my system (Win XP x64)

Even Lib32/Lib64 files did not help

I guess your system is Win 7,right ?

Best,
Sedat

[velmarin]

George, thank you very much for the test.

Sedat, here you have an executable with Visual C + +, and no instructions POPCNT,
lost some speed, how about 10 for 100, but will not be serious.

Would be interested in if it works.
http://www.mediafire.com/?95v574a0kj66d5u

Thanks again George, very grateful.

[geots]

Thanks for your interesting testing dear George

It seems,Bouquet 1.4 is a very strong engine...

But unfortunately,i could not run Bouquet 1.4 on my system (Win XP x64)

Even Lib32/Lib64 files did not help

I guess your system is Win 7,right ?

Best,
Sedat

Yes, it is Intel i5 running on Win7. But I know someone who is testing it on xp if I am not mistaken. But I may be wrong. I will check tomorrow and let you know.


Best,

george

[geots]

George, thank you very much for the test.

Sedat, here you have an executable with Visual C + +, and no instructions POPCNT,
lost some speed, how about 10 for 100, but will not be serious.

Would be interested in if it works.
http://www.mediafire.com/?95v574a0kj66d5u

Thanks again George, very grateful.

I am not thru yet. Before I head to get some sleep in an hour or so, I am going to set up another match with 1.4 and an engine I have not decided on yet. I was looking at Quazar- which is rated 35 or 40 points ahead of Bouquet 1.2. But even that is a waste of good time. I ran a short 20 game 3 minute per game match between Quazar and Bouquet 1.4 last night, and it was like men against boys. Bouquet literally beat his brains out. I will find something, however- it is just a matter of looking and thinking a bit.


I can't remember if I told you this, but if I did I will be telling you again: I am very proud of you.


Best,

george

[geots]

George, thank you very much for the test.

Sedat, here you have an executable with Visual C + +, and no instructions POPCNT,
lost some speed, how about 10 for 100, but will not be serious.

Would be interested in if it works.
http://www.mediafire.com/?95v574a0kj66d5u

Thanks again George, very grateful.

Velmarin, I am going to move up another 34 or 35 points higher than Zappa Mexico II, and put 1.4 up against spark-1.0. This engine may be problematic for you, because it has some better speed than Zappa. But there is only one way to find out. Win or lose, you got one hell of a good engine. We shall see; results will be posted tonight- CDT USA.


george

[Dr.Wael Deeb]

Thanks for your interesting testing dear George

It seems,Bouquet 1.4 is a very strong engine...

But unfortunately,i could not run Bouquet 1.4 on my system (Win XP x64)

Even Lib32/Lib64 files did not help

I guess your system is Win 7,right ?

Best,
Sedat

Same problem here Sedat....

[Uri Blass]

George, thank you very much for the test.

Sedat, here you have an executable with Visual C + +, and no instructions POPCNT,
lost some speed, how about 10 for 100, but will not be serious.

Would be interested in if it works.
http://www.mediafire.com/?95v574a0kj66d5u

Thanks again George, very grateful.

I am not thru yet. Before I head to get some sleep in an hour or so, I am going to set up another match with 1.4 and an engine I have not decided on yet. I was looking at Quazar- which is rated 35 or 40 points ahead of Bouquet 1.2. But even that is a waste of good time. I ran a short 20 game 3 minute per game match between Quazar and Bouquet 1.4 last night, and it was like men against boys. Bouquet literally beat his brains out. I will find something, however- it is just a matter of looking and thinking a bit.

george

I think that you probably need a program with ccrl rating of 3000-3100

[Sedat Canbaz]

George, thank you very much for the test.

Sedat, here you have an executable with Visual C + +, and no instructions POPCNT,
lost some speed, how about 10 for 100, but will not be serious.

Would be interested in if it works.
http://www.mediafire.com/?95v574a0kj66d5u

Thanks again George, very grateful.

Thanks,but the same problem continues as before



And i hope there will be a new Bouquet release for Win XP systems too

Greetings,
Sedat