Homepage

superchessengine.com

 Engine testing condition

Cpu

Memory

total

Memory

hash

EGTB

(3-4-5??)

EGTB

cache

Time

Tournament

type

Top6.pgn

Results:

Homepage

superchessengine.com

 Engine testing condition

Cpu

Memory

total

Memory

hash

EGTB

(3-4-5??)

EGTB

cache

Time

Tournament

type

Results:

I tested the New Glaurung SMP 64 on my Xeon vs 5 commercial and vs 5 top Free UCI engines

Thanks to Tord for the new Glaurung 64 bits (up to 8 cpu and maybe more ??)
It was was very fast (over 8000 kn/s on some games)
Of course this is not a Rybka 2.3.2a smp Killer . it lost vs DS11, Zappa Mexico,Naum ( the top commercial engines behind Rybka ) and beat Fritz 10 ( a good commercial engine )

Now begin the interesting test vs Top Free UCI engines, Glaurung won over all top Free UCI in this test (Including Free Rybka, Fruit 2.3.3J, Toga, Belka and Spike

At least on my big hardware (8 CPU ) A new king for Free UCI engine is born

Homepage

superchessengine.com

 Engine testing condition

Cpu

Memory

total

Memory

hash

EGTB

(3-4-5??)

EGTB

cache

Time

Tournament

type

Results:

Zappa Scale better , however Rybka 2.3.2 a SMP 64 is still the best chess engine out there even on 8 cpu

around 50 elo after 50 games the difference is not so big on 8 cpu .
A good tournament book could overcome the 50 elo difference especially when it is tuned versus against this engine . Zappa Team probably tuned their white and black book versus Rybka . Their advantage was that almost the same Rybka they tested played against them ...
A.Cozzie did a very job improving his zappa & Erdo made a very good book tuned vs Rybka + a little bit luck and we had Zappa Mexico team victory at Mexico (congrats to them one more time )

Zappa Victory vs Rybka at Mexico was probably 50 % engine + Xeon + 25 % book + 25 % luck

Homepage

superchessengine.com

 Engine testing condition

Cpu

Memory

total

Memory

hash

EGTB

(3-4-5??)

EGTB

cache

Time

Tournament

type

>5 mn

Results:

I just finished the test for Glaurung & Belka in a 4 Round Robin vs some top Chess engines

For this one I decide to use Fritz 10 which is a very good real tournament book (not draw book)

1st - With no Surprise Rybka of course Finished 1st ... a little surprise is that Rybka didn t kick Fruit and Toga Ass ...

2nd - is DS11 who did very good except on this tournament, It even kill Naum smp (3.5/4) , bad think is of course vs Rybka 0.5/4 and Zappa

3rd - Zappa Mexico just 0.5 behind DS11

4th - New Glaurung was impressive on 8 cpu , Maybe the new Free UCI engine King at least on 8 cpu (Wait and See )

5 - Naum did well vs Zappa (50 % ) However it lost too much points vs Rybka and DS11 on this test

6 - This Fruit Beta is for sure stronger than Fruit 2.3.1 (vs others chess engines ) , It was the only to beat Rybka on 4 games match

7 - Spike Draws a lot of games

8 - Toga did so so ... not too bad not too good ..

9 - Fritz 10 ... This is a surprise to see it so low because the test was with his own book supposed to be tuned for him !!!

10 - Belka played too fast , I expected him to be in the middle of this test , Strangely it did very bad only 2/36 !!! Maybe something gone wrong ....
especially in the end of game where it was killed
 

Homepage

superchessengine.com

 Engine testing condition

Cpu

Memory

total

Memory

hash

EGTB

(3-4-5??)

EGTB

cache

Time

Tournament

type

>360s

Results:

For Blitz

Zappa Scale better , however Rybka 2.3.2 a SMP 64 is still the best chess engine out there even on 8 cpu

Mama Milla,  Naum was too optimist versus Rybka , Naum show positive score when Rybka see the opposite, Naum is very strong (probably one of the top 3-5 chess engine ) but this time was punish by Rybka

Homepage

superchessengine.com

 Engine testing condition

Cpu

Memory

total

Memory

hash

EGTB

(3-4-5??)

EGTB

cache

Time

Tournament

type

Results:

1- Glaurung SMP 64 bits and Zappa Mexico tied for 1st with 11.5/15 and both undefaited . Very good for Glaurung !!!

3- Fre Rybka 64 bits - still very strong

4-Fritz 10 did very good only 1 loss

5- Naum 2.2 SMP , something strange 4 losses in the row !!! maybe a bug... don t know ... when I retarted the computer 4 wins in the row !!!

Main Fact in this tournament :

DS 11 did so so (not bad , not good)
Fruit beta is very strong and finished in the top 10...
Tested Crafty smp 8 cpu compiled by Jim, did very good on my Xeon
Deep Pharaon Crashed ... 3-4 times

Homepage

superchessengine.com

 Engine testing condition

Cpu

Memory

total

Memory

hash

EGTB

(3-4-5??)

EGTB

cache

Time

Tournament

type

Results:

Deep Shredder 11 did better this time . However on 8 cpu it is a little bit under Zappa Mexico level

It score 45 % vs Naum 2.2 SMP (Naum 2.2 SMP belongs definively to the top 3-5 chess engine )

Like I folows some testers test too

On 1 cpu

Shredder is a lot more stronger than Zappa Mexico
Shredder is a lot more stronger than Naum 2.2


On 4 CPU

Zappa Mexico Stronger than Deep Shredder
Deep Shredder and Naum approximatively same level


on 8 CPU

Zappa Mexico a lot more stronger than Deep Shredder
Naum 2.2 MP stronger than Deep shredder

 

Homepage

superchessengine.com

 Engine testing condition

Cpu

Memory

total

Memory

hash

EGTB

(3-4-5??)

EGTB

cache

Time

Tournament

type

Results:

DS 11 is  in big trouble vs Zappa And Naum (all play with 8 cpu)

score only 22.5 % vs Zappa Mexico and 35% vs Naum 2.2 SMP

 DS is a lot more comfortable vs Top free UCI  around 65 % vs Glaurung and Fruit 2.3.3J Beta
after a bad begining DS 11 fought back and has a positive record vs Free Rybka 1.0 64 bits

Homepage

superchessengine.com

 Engine testing condition

Cpu

Memory

total

Memory

hash

EGTB

(3-4-5??)

EGTB

cache

Time

Tournament

type

512Mo(SMP)

>240s

Results:

This version 2.3.3 J is well tuned vs Free Rybka and Toga ... The 2 top free UCI engine the last 2 years

It  managed to beat all free Top UCI engine (  Rybka, Toga, Pharaon ,Naum, Spike ,Glaurung...) in a 20 games match except Strelka.

It draw vs Fruit 2.3.1 .

It has a lot of difficulties vs Strelka 1.8 (a lot of controversy around Strelka ,illegal  Clone or not?I don't know & I m not the one to decide)

It was not a surprise to see Fruit lost againstcommercial Zappa Mexico 8 CPU

I m running some  test for  Fruit beta... More version will be test ...

This one did well versus strong Free UCI engine

Homepage

superchessengine.com

 Engine testing condition

Cpu

Memory

total

Memory

hash

EGTB

(3-4-5??)

EGTB

cache

Time

Tournament

type

512Mo(SMP)

>360s

Results:

Zappa Mexico is Very impressive on 8 CPU (just look at the match RybkaSMP vs Zappa Mexico!!)

That s why I decide to test it on my 2Xeon 5365 (8CPU 3ghz) 

Zappa was able to score 74.5 % vs 10 top Free uci Engine!!! (and this is just The 32 bits version)

It managed to score positive record vs all opponent (only Strelka draw it, I was alittle bit surprise )

Toga,Strelka and Glaurung SMP (it was on 4 CPU) did very well vs Zappa Mexico (A top 2 commercial engine )

This is the first Time I saw a Top Commercial engine other than Rybka that could beat easily the top 10 free UCI engine ...

It was a big surprise to see Zappa Mexico score 80 % vs Free Rybka 1.0 (A top 3 free engine , maybe the best free engine). It gave a chess lesson to others free UCI engine (95 % vs Pharaon !!! 80 % vs Fruit 2.3.1, 80% vs Spike 1.2 ...)