Regarding SSDF.

[Nimzovik]

Hmmm... Seems like non deep programs are not being tested -Fritz 12 etc. I wonder how the IPP family would fare.

[Stephen Ham]

Hmmm... Seems like non deep programs are not being tested -Fritz 12 etc. I wonder how the IPP family would fare.

Non-"deep" engines such as Fritz 12 have been tested.

http://ssdf.bosjo.net/list.htm

Steve

[Nimzovik]

Indeed ..however on Athalon processors. What is the point of that? Note the rybka 2.3 vs Naum on Athalon. What a waste of time.

[Dann Corbit]

Indeed ..however on Athalon processors. What is the point of that? Note the rybka 2.3 vs Naum on Athalon. What a waste of time.

The programs are extremely well calibrated on Athlon processors (having many thousands of measurements performed).

In order to switch to Intel, they would have to start from scratch. It would take many months before the data measurements gave as much information as the current machine measurements.

Of course, if you were to buy high end Intel machines for the SSDF volunteers, they might well give it a go.

[Nimzovik]

Very true indeed. I am not that rich however. I guess what I am saying is that is it really fair to say what the Athalon processors indicate is their true rating? I just personally am saying I would like to know their (the other engines) accurate perfromance on equivalent equipment. This does not mean I am not appreciative of their work however such as it is.....

[Uri Blass]

Indeed ..however on Athalon processors. What is the point of that? Note the rybka 2.3 vs Naum on Athalon. What a waste of time.

the athlon is no excuse for poor result

From the ssdf list in dos format.

7 Deep Rybka 3 256MB Athlon 1200 MHz 3084 39 -38 332 58%

18 Fritz 12 256MB A1200 MHz 2939 38 -36 370 64% 2835

You can see that Fritz12 is 145 elo weaker than an old version of Rybka on the same athlon(and Deep Rybka3 cannot earn from parallel search on that hardware)

It is going to be a waste of time to test Fritz12 on a better hardware that can use more than one core because it means that part of the cores are going to do nothing.

Uri

[Dr.Wael Deeb]

Indeed ..however on Athalon processors. What is the point of that? Note the rybka 2.3 vs Naum on Athalon. What a waste of time.

the athlon is no excuse for poor result

From the ssdf list in dos format.

7 Deep Rybka 3 256MB Athlon 1200 MHz 3084 39 -38 332 58%

18 Fritz 12 256MB A1200 MHz 2939 38 -36 370 64% 2835

You can see that Fritz12 is 145 elo weaker than an old version of Rybka on the same athlon(and Deep Rybka3 cannot earn from parallel search on that hardware)

It is going to be a waste of time to test Fritz12 on a better hardware that can use more than one core because it means that part of the cores are going to do nothing.

Uri

Totally false statement....are you saying that testing Fritz 12 on an Athlon processor is the same as testing it on one core i7

Ever considered the the much fastest architecture of a particular core compared to an old one

[Nimzovik]

Yes...... That was my thoughts as well....... I was wondering How Fritz 12 would compare on equivalent equipment. I 'm sure it won't do that well compared to other programs, however I just want a more accurate figure. BTW does Fritz 12 do any type of position better than the top programs? Hard to answer that one. I am not looking for hardcore test data. Just gut guesses.

[Nimzovik]

Interesting is the absence of Rybka 4 (not to mention Rybka 3+ .) No doubt it will be added when the sampling is large enuff.

[Dr.Wael Deeb]

Yes...... That was my thoughts as well....... I was wondering How Fritz 12 would compare on equivalent equipment. I 'm sure it won't do that well compared to other programs, however I just want a more accurate figure. BTW does Fritz 12 do any type of position better than the top programs? Hard to answer that one. I am not looking for hardcore test data. Just gut guesses.

I am testing Fritz 12 right now on my overclocked Q9650 but it's early days to give any estimation about it's performance....
Dr.D