Multi Processor Chess Benchmark!

[JPQY]

Test2 is with a Multi Processor Chess Engine!

Download the program Arena+Test2 with this link: (Faster way)

http://www.cranox.com/Software/Arena/Arenatest2.zip

1.Unrar only the map "Arena"
2.Start Arena
3.You will see a window pop-up
4.Click two times on "Administator" (This is the FEN-code)
5.Pop-up screen close and the Test position come on the chessbord
6.Go to "Levels" and select "Infinite"
7.Press on Space-Bar

Computer begins to calculate and you have to wait till you see in the 5th row "+M13" means Mate in 13 moves! not before,before you can see a positieve score,then +M21,+M16,+M15,+M14..

When you see the "+M13" you can stop calculating by clicking on "Stop"
You can make a copy from your screen or go to "Position" and select "Copy Analyses to Clipbord" and paste your result in a new message on the forum,or in e-mail!Thanks!

Regards,JP.

If you have any question,feel free to ask me or you can find me on MSN: jpqy@hotmail.com

Hope your test go well

Regards,Jean-Paul.

Standings:
1. Core 2 Duo E6700 at 3.33Ghz (+M13) Depth: 19 in 01m29s!! (143Mb HT)
2. Core 2 Duo E6700 at 3.2Ghz (+M13) Depth: 23 in 03m06s (143Mb HT)
3. AMD X2 5000+ at 3.0Ghz (+M13) Depth: 23 in 03m34s (143Mb HT)
4. AMD Opteron 165 at 2.85Ghz (+M13) Depth: 22 in 04m56s (143Mb HT)
5. Core 2 Duo E6600 at 2.4Ghz (+M13) Depth: 23 in 04m58s (143Mb HT)
6. Core 2 Duo E6600 at 3.2Ghz (+M13) Depth 24 in 05m17s (143Mb HT)
7. AMD Athlon X2 4200 (+M13) Depth: 24 in 05m32s (256mb HT)
8. Core 2 Duo E6700 at 2.66Ghz (+M13) Depth: 25 in 06m38s (143Mb HT)
9. Quad Opteron 870 at 2.0Ghz (+M13) Depth: 25 in 07m03s (144Mb HT)
10. AMD Turion 64 X2 at 1.6Ghz (+M13) Depth: 25 in 15m52s (1Gb HT)
11. AMD 3400+ at 2.4Ghz (+M13) Depth 27 in 32m57s (79Mb HT)
12. AMD 3800+ at 2.5Ghz (+M13) Depth 28 in 38m21s
13. Pentium M755 at 2Ghz (+M13) Depth 28 in 48m31s

Where are the X6800 and the Kentsfield to be my new number one!

[Qkjhfhaiguihfma]

why the :banana::banana::banana::banana: did autoarenatest start moving my mouse around and doing weird :banana::banana::banana::banana: before i killed it?

if this was some kind of spyware, you will pay dearly.

[JPQY]

Hi,i don't understand people who thinks it is spyware,i put alot of time to make a good test,i'am 40 years and sorry i have no time for this!
If your mouse do strange than you tell me what goes wrong,other people have done before and it work very well!
I think it have to make with the size of your screen,i have make this macro on a 19" screen,so if you have done the test on a 17" then is it possible that the macro do wrong on your screen!
So if you still want to try the test i can send you a message how to do without the macro!

Thanks,JP

[[cTx]Philosophy]

Its a macro dude chill, which is basically your computer controlling your pc to do a certain task man, I doubt its spyware, thats just what macros do, they can control your mouse and kb and such.. click type move etc etc etc..

[JPQY]

Ok,to many problems with macro...here the same test but don't use the macro!and follow the steps...very simple and a few seconds to set-up!!
This is No Spyware,No Trojan,No Virus,No Spam...It is a Full Working Chess Program with a Top Chess Engine!!
This is my hobby,and i just want to make a list with your time results!

Download the program Arena+Test2 with this link:

http://www.cranox.com/Software/Arena/Arenatest2.zip

1.Unrar only the map "Arena"
2.Start Arena
3.You will see a window pop-up
4.Click two times on "Administator" (This is the FEN-code)
5.Pop-up screen close and the Test position come on the chessbord
6.Go to "Levels" and select "Infinite"
7.Press on Space-Bar

Computer begins to calculate and you have to wait till you see in the 5th row "+M13" means Mate in 13 moves! not before,before you can see a positieve score,then +M21,+M16,+M15,+M14..

When you see the "+M13" you can stop calculating by clicking on "Stop"
You can make a copy from your screen or go to "Position" and select "Copy Analyses to Clipbord" and paste your result in a new message on the forum!Thanks!

Regards,JP.

[JPQY]

This is a new result from a Core2Duo E6700!! (click on it to zoom out)

http://img337.imageshack.us/img337/4688/chessfd0.jpg

Who do better,give it a try,just a few seconds to set-up the test!!
The best Multi Processor Chess Engine from the world!!

JP.

PS. :clap: Come on Xtreme people let me see what you get!! :clap:

[pH(x)]

Will this macro work with HyperThreading, JPQY?

[JPQY]

Hi pH(x) ,it works on every computer,but don't use it if you don't have a 19" monitor,because it is made on my computer and find out that many people have problem with it,so it is better to take the last test without marco!
My monitor is a Samsung SyncMaster 913n.
Jean-Paul.

[SMa]

why the :banana::banana::banana::banana: did autoarenatest start moving my mouse around and doing weird :banana::banana::banana::banana: before i killed it?

if this was some kind of spyware, you will pay dearly.
It's not spyware. In fact it's very trustworthy.
We tested it already on a Belgium forum and no problems occured.

[JPQY]

A new number one!! E6700 @ 333x10 1.4vCore

http://www.systekonline.com/upload/files/202/Public%20Folder/Test2NickE6700.png

Results for Mate in 13!:
1.Core 2 Duo E6700 at 3.33Ghz (+M13) Depth: 19 in 01m29s!! (143Mb HT)
2.Core 2 Duo E6700 at 3.2Ghz (+M13) Depth: 23 in 03m06s (143Mb HT)
3.Core 2 Duo E6600 at 2.4Ghz (+M13) Depth: 23 in 04m58s (143Mb HT)
4.Core 2 Duo E6700 at 2.66Ghz (+M13) Depth: 25 in 06m38s (143Mb HT)
5.AMD Athlon X2 4200 (+M13) Depth: 24 in 05m32s (256mb HT)
6.AMD Turion 64 X2 at 1.6Ghz (+13) Depth: 25 in 15m52s
7.AMD 3400+ at 2.4Ghz (+M13) Depth 27 in 32m57s (79Mb HT)
8.AMD 3800+ at 2.5Ghz (+M13) Depth 28 in 38m21s
9.Pentium M755 at 2Ghz (+M13) Depth 28 in 48m31s

Who do better,give it a try,just a few seconds to set-up the test!!
The best Multi Processor Chess Engine from the world!!

JP.

PS. Come on Xtreme people let me see what you get!!

And Thanks SMa!!! JP.

[colfin22]

Mine:
http://img45.imageshack.us/img45/5171/benchmarkll0.th.jpg (http://img45.imageshack.us/my.php?image=benchmarkll0.jpg)

Edit: cpu-z should read like this:
http://img165.imageshack.us/img165/5545/cpuzhf1.th.jpg (http://img165.imageshack.us/my.php?image=cpuzhf1.jpg)
i'm on a beta bios & am having issues with it. c1e somehow re-enabled itself.

[alpha754293]

JP:

I did get a chance to run your benchmarks (and also the other guy's).

For multi-processor "mystery system #1", try and see if you can guess what my system configuration is with this:

http://www.xtremesystems.org/forums/attachment.php?attachmentid=51517&stc=1&d=1158897256

I don't understand it. I ran the same benchmark twice and I get different results. *shrug*

[alpha754293]

There are a couple of problems that I've found with your program.

a) The CPU on my "mystery system 2" is at least 11% faster when compared to the CPU on my "mystery system 1". Yet, in the run that I just did, it took nearly three times as long to get to +M13. That is inconsistent with the benchmarks results that I have already posted (taken with Sandra 2005) and also with the 17 queens program that the other person wrote.

b) The program stops being multi-processor/multi-threaded after about 1 m 40 s, or when it passes "Depth 20" and only uses one processor/one-thread at that point which dramatically slows down.

c) Comparing to the other 17 queens program; that one shows a consistent scaling per the number of CPUs/threads engaged. Your program does not run the same way twice.

All of those things invalidate any of the results you have gotten thus far, because the results are inconsistent.

Please have a look at it.

http://www.xtremesystems.org/forums/attachment.php?attachmentid=51537&stc=1&d=1158932893

I re-ran it just to make sure that the times were consistent. They're not! (and this is running it twice back-to-back on the same system)

http://www.xtremesystems.org/forums/attachment.php?attachmentid=51538&stc=1&d=1158933162

Oops, I took the screen shot wrong, but you get the idea. 11 minutes vs 15 minutes on the same system. 0.o???

[VictorWang]

I think it's easy for QX6700 :)

3.5G , 900+

[I_G]

Trying it now...

[Face]

Umm..Hi guys.

I hate being the party crasher but.. probably without your knowledge you have been using pirated software that was downloaded into your computer.
software that costs $73.
file "Rybka v2.1c.mp.w32.exe" is in fact the the Rybka chess engine known to be the stronget chess engine available today, and it is being sold commercially. (see http://rybkachess.com)

so I think it would be morally&legally right to remove the link(s) and ask the host to remove the file from the server.

[JPQY]

My list begin to grow nice now!
But i get people who are dissapointed with there result if they see that there faster computer system stays lower then a other.
But like i have said many times,it depends on your settings,and at some reason you can get bottleneck situation in your system and it's not always with the highest settings that you get your fastest computer.
Today i get a new result from a E6400!! and look he take the second place,just great job he do,very good settings he have!
He give me also his Ram setting.So i think it's very important to find out witch settings are the best for your system!

Grtz,Jean-Paul.

Results for Mate in 13 :
1. Core 2 Duo E6700 at 3.33Ghz (+M13) Depth: 19 in 01m29s!! (143Mb HT)
2. Core 2 Duo E6400 at 3.6Ghz (+M13) Depth: 20 in 01m50s!! (…Mb HT) Patriot 667 LLK @ 1125 mhz 5-5-5-5 2.15v
3. Core 2 Duo X6800 at 4.4Ghz (+M13) Depth: 22 in 02m11s (…Mb HT) on Vista
4. Core 2 Duo X6800 at 3.6Ghz (+M13) Depth: 23 in 02m51s (143Mb HT)
5. Core 2 Duo E6700 at 3.35Ghz (+M13) Depth: 22 in 02m53s (143Mb HT)
6. Core 2 Duo E6700 at 3.2Ghz (+M13) Depth: 23 in 03m06s (143Mb HT)
7. Core 2 Duo X6800 at 4.5Ghz (+M13) Depth: 25 in 03m17s (…Mb HT)
8. Kentsfield E6700ES at 2.93Ghz (+M13) Depth 23 in 3m18s (143Mb HT)
9. Core 2 Duo E6600 at 3.0Ghz (+M13) Depth: 23 in 03m21s (143Mb HT)
10. AMD X2 5000+ at 3.0Ghz (+M13) Depth: 23 in 03m34s (143Mb HT)
11. Kentsfield E6700ES at 2.8Ghz (+M13) Depth 23 in 3m54s (143Mb HT)
12. Core 2 Duo X6800 at 4.4Ghz (+M13) Depth: 25 in 04m27s (…Mb HT) on XP
13. AMD Opteron 165 at 2.85Ghz (+M13) Depth: 22 in 04m56s (143Mb HT)
14. Core 2 Duo E6600 at 2.4Ghz (+M13) Depth: 23 in 04m58s (143Mb HT)
15. Core 2 Duo E6600 at 3.2Ghz (+M13) Depth 24 in 05m17s (143Mb HT)
16. AMD Athlon X2 4200+ (+M13) Depth: 24 in 05m32s (256Mb HT)
17. Kentsfield E6700ES at 3.73Ghz (+M13) Depth: 25 in 06m29s (540Mb HT)
18. Core 2 Duo E6700 at 2.66Ghz (+M13) Depth: 25 in 06m38s (143Mb HT)
19. Quad Opteron 870 at 2.0Ghz (+M13) Depth: 25 in 07m03s (144Mb HT)
20. Core 2 Duo E6300 at 3.3Ghz (+M13) Depth 26 in 07m59s (144Mb HT)
21. Core 2 Duo E6300 at 3.2Ghz (+M13) Depth 24 in 08m39s (143Mb HT)
22. Kentsfield E6700ES at 3.73Ghz (+M13) Depth 27 in 9m23s (143Mb HT)
23. AMD Athlon X2 4800+ at 2.4Ghz (+M13) Depth 25 in 9m32s (143Mb HT)
24. AMD Turion 64 X2 at 1.6Ghz (+M13) Depth: 25 in 15m52s (1Gb HT)
25. AMD 3400+ at 2.4Ghz (+M13) Depth 27 in 32m57s (79Mb HT)
26. AMD 3800+ at 2.5Ghz (+M13) Depth 28 in 38m21s
27. Pentium M755 at 2Ghz (+M13) Depth 28 in 48m31s

Here i put a list with the average nodes/second from the first second he begins to calculate:

Results for Mate in 13 : Average Nodes/Second
1. Quad Opteron 870 at 2.0Ghz (+M13) Depth: 25 854021 n/s
2. Core 2 Duo X6800 at 4.4Ghz (+M13) Depth: 22 651911 n/s
3. Kentsfield E6700ES at 3.73Ghz (+M13) Depth: 25 578174 n/s
4. Core 2 Duo X6800 at 4.4Ghz (+M13) Depth: 25 570938 n/s
5. Core 2 Duo E6300 at 3.3Ghz (+M13) Depth 26 504206 n/s
6. Core 2 Duo X6800 at 4.5Ghz (+M13) Depth: 25 485151 n/s
7. Core 2 Duo X6800 at 3.6Ghz (+M13) Depth: 23 476892 n/s
8. Core 2 Duo E6700 at 3.33Ghz (+M13) Depth: 19 455385 n/s
9. Core 2 Duo E6700 at 3.5Ghz (+M13) Depth: 25 451134 n/s
10. Core 2 Duo E6600 at 3.0Ghz (+M13) Depth: 23 420745 n/s
11. AMD X2 5000+ at 3.0Ghz (+M13) Depth 23 369592 n/s
12. AMD Opteron 165 at 2.85Ghz (+M13) Depth 22 348880 n/s
13. Core 2 Duo E6700 at 3.2Ghz (+M13) Depth: 23 331898 n/s
14. AMD X2 4800+ at 2.4Ghz (+M13) Depth 25 272853 n/s
15. Core 2 Duo E6700 at 2.66Ghz (+M13) Depth 25 247532 n/s
16. Core 2 Duo E6600 at 2.4Ghz (+M13) Depth: 23 233718 n/s
17. AMD Turion 64 X2 at 1.6Ghz (+M13) Depth: 25 189016 n/s
18. AMD 3800+ at 2.5Ghz (+M13) Depth 28 161099 n/s
19. AMD 3400+ at 2.4Ghz (+M13) Depth 27 135623 n/s
20. Pentium M755 at 2Ghz (+M13) Depth 28 126985 n/s

[robhalford88]

Is this any good?
FEN: R7/1b3r1p/2pnk3/2Qp1p2/3BB3/6RP/P1P3PK/8 b - - 1 1

Rybka v2.1c.mp.w32:
2 00:00 64 4.096 +8.80 Bf3xe4 Bb7xa8
3 00:00 144 4.608 +8.80 Bf3xe4 Bb7xa8 Be4d3
4 00:00 560 17.920 +8.23 Bf3xe4 Bb7xa8 Be4d3 Nd6e4
5 00:00 1.008 32.256 +8.07 Bf3xe4 Bb7xa8 Be4d3 Nd6e4 Bd3xe4 f5xe4 Rg3g7
6 00:00 4.192 89.429 +8.41 Bf3xe4 Bb7xa8 Be4d3 Ba8b7 Rg3g7 Rf7xg7 Bd4xg7 h7h5
7 00:00 10.960 119.394 +8.79 Bf3xe4 Bb7xa8 Be4d3 Nd6e4 Bd3xe4 f5xe4 Rg3g7 Rf7xg7 Bd4xg7 Ba8b7
8 00:00 11.936 77.850 +8.79 Bf3xe4 Bb7xa8 Be4d3 Nd6e4 Bd3xe4 f5xe4 Rg3g7 Rf7xg7 Bd4xg7 Ba8b7 Qc5f8 Bb7a6
9 00:00 25.696 120.149 +8.95 Bf3xe4 Bb7xa8 Be4d3 Nd6e4 Bd3xe4 f5xe4 Rg3g7 Rf7xg7 Bd4xg7 Ba8b7 Qc5f8 Bb7a6
10 00:00 35.792 137.785 +8.95 Bf3xe4 Bb7xa8 Be4d3 Nd6e4 Bd3xe4 f5xe4 Rg3g7 Rf7xg7 Bd4xg7 Ba8b7 Qc5f8 Bb7a6 Qf8f6+ Ke6d7
11 00:00 130.384 152.412 +9.57 Bf3xe4 Bb7xa8 Be4d3 Ba8b7 Rg3g7 Rf7xg7 Bd4xg7 h7h5 Qc5e3+ Ke6f7 Qe3e5 Nd6e8 Qe5xf5+ Kf7xg7
12 00:01 215.296 195.793 +9.66 Bf3xe4 Bb7xa8 Be4d3 Ba8b7 Rg3g7 Rf7xg7 Bd4xg7 h7h5 Qc5e3+ Ke6d7 Qe3e5 c6c5 Bd3xf5+ Nd6xf5
13 00:02 494.384 238.122 +9.81 Bf3xe4 Bb7xa8 Be4d3 Ba8b7 Rg3g7 Rf7xg7 Bd4xg7 f5f4 Bd3xh7 Ke6d7 Qc5b4 Bb7a6 Qb4xf4 Ba6b5
14 00:03 1.059.520 277.693 +11.10 Bf3xe4 Bb7xa8 Be4d3 Ba8b7 Rg3g7 Rf7xg7 Bd4xg7 h7h6 Bg7f8 Nd6c8 Qc5e3+ Ke6d7 Bd3xf5+ Kd7d8
15 00:05 1.912.048 322.932 +11.58 Bf3xe4 Bb7xa8 Be4d3 Ba8b7 Rg3g7 Rf7xg7 Bd4xg7 Nd6e4 Bd3xe4 f5xe4 Qc5b6 Bb7a8 Qb6a7 h7h5
16 00:09 3.301.248 357.571 +12.15 Bf3xe4 Bb7xa8 Be4d3 Ba8b7 Rg3g7 Rf7xg7 Bd4xg7 Nd6e4 Bd3xe4 f5xe4 Qc5b6 Bb7a8 Qb6a7 e4e3
17 00:18 5.248.064 291.211 +12.15 Bf3xe4 Bb7xa8 Be4d3 Ba8b7 Rg3g7 Rf7xg7 Bd4xg7 h7h5 Qc5b6 f5f4 Bg7f8 Bb7a8 Bf8xd6 Ke6xd6
18 00:26 8.312.896 321.599 +12.31 Bf3xe4 Bb7xa8 Be4d3 Ba8b7 Rg3g7 Rf7xg7 Bd4xg7 h7h5 Qc5b6 f5f4 Bg7f8 Bb7a8 Bf8xd6 Ke6xd6
19 00:46 16.053.328 350.100 +12.87 Bf3xe4 Bb7xa8 Be4d3 Ba8b7 Rg3g7 Rf7xg7 Bd4xg7 h7h5 Qc5b6 f5f4 Bg7f8 Ke6d7 Bf8xd6 Kd7xd6
20 04:50 64.466.592 226.839 +M20 Bf3xe4 Bb7xa8 Be4d3 Ba8b7 Rg3g7 Rf7xg7 Bd4xg7 Bb7a8 Qc5e3+ Ke6d7 Qe3a7+ Kd7e6 Qa7xa8 h7h6 Bg7xh6 c6c5 Qa8g8+ Ke6d7 Qg8xd5 Kd7e7 Qd5xc5 Ke7f7 Qc5xd6 f5f4 Bd3g6+
21 05:20 71.692.704 229.113 +M19 Bf3xe4 Bb7xa8 Be4d3 Ba8b7 Rg3g7 Rf7xg7 Bd4xg7 f5f4 Bd3xh7 f4f3 g2xf3 Bb7a8 Qc5a7 Ba8b7 Bg7f8 Ke6f7 Bf8xd6 Kf7e6 Qa7b8 Bb7c8
22 05:45 80.315.136 237.780 +M17 Bf3xe4 Bb7xa8 Be4d3 Ba8b7 Rg3g7 Rf7xg7 Bd4xg7 f5f4 Bd3xh7 f4f3 g2xf3 Bb7a8 Qc5a7 Ba8b7 Bg7f8 Ke6d7 Qa7b8 Nd6c8 Qb8xb7+ Kd7d8 Qb7xc6
23 08:10 132.240.736 275.784 +M13 Bf3xe4 Bb7xa8 Be4d3 Ba8b7 Rg3g7 Rf7xg7 Bd4xg7 f5f4 Bd3xh7 f4f3 g2g4 Ke6d7 Bg7e5 Nd6e8 Bh7f5+ Kd7d8 Qc5b6+ Kd8e7 Qb6xb7+ Ke7f8 Bf5g6 Ne8c7 Qb7xc7

This is a dual Opteron 252 with 4gb of ram.

[alpha754293]

Is this any good?
23 08:10 132.240.736 275.784 +M13 Bf3xe4 Bb7xa8 Be4d3 Ba8b7 Rg3g7 Rf7xg7 Bd4xg7 f5f4 Bd3xh7 f4f3 g2g4 Ke6d7 Bg7e5 Nd6e8 Bh7f5+ Kd7d8 Qc5b6+ Kd8e7 Qb6xb7+ Ke7f8 Bf5g6 Ne8c7 Qb7xc7

This is a dual Opteron 252 with 4gb of ram.

The first number is the "depth" - i.e. how many levels or iterations of searches it has to go through to find the a way to win (based on the given criteria of checkmate in 13 moves).

The time is, well, the time.

Unlike most other benchmarks, don't take the "time" to be absolute because if you run the benchmark a couple of times, it is highly unlikely that you will get the same time twice.

If you look at the 275.784, that means 275,784 nodes/sec. That is the rate at which the program running on your system is able to find a solution to the problem.

JP and I have talked about revamping his list so that it sorts the results by nodes instead BECAUSE the runtime isn't constant. There is no direct correlation between system performance/specifications and the run time. However, there is a direct correlation between the computational rate (i.e. node/sec) and the system specs/performance.

It took me a while of running his program to actually understand what most of it means.

It's an interesting one - it's not quite as straightforward as the "common standard set" I think because people are too "conditioned" to use time as the absolute measure for performance. For a highly nonlinear problem such as a chess engine; using time as the benchmark metric breaks down.

(P.S. Yes, I am presuming that the chess problem is a highly nonlinear problem. I don't know enough about chess engines or programming to know either way.)