Real PR (performance rating) of the processors to be tested

 

I have noticed that PR calculated by the tester does not reflect the real situation. Particularly, Intel T2400 shows PR56 on one core and PR70 on the other, while AMD Opteron 6128 shows PR25 to PR49 on different cores. It would seem that the former should be ahead of the latter. In reality, my Expert Advisor has a 50% advantage over the latter! The latter has a smaller but comparatively big advantage, about 15% on MACD.

I suggest to publish the results of testing MACD Sample with default parameters on EURUSD M1 history 2010.01.01-2010.08.01.

For the sake of purity of the experiment, log out of the terminal before testing, then log in again.

Give the name of the processor as it is shown in the Agents tab and the line with the test result from the Journal.

Here are my two results:

Processor: Intel T2400 @ 1.83GHz, 3062MB, PR70
EURUSD,M1: 7143130 ticks (212232 bars) generated within 26468 ms (total bars in history 576660)
Time: 26.5 seconds

Processor: AMD Opteron 6128, 9598MB, PR49
EURUSD,M1: 7143130 ticks (212232 bars) generated within 22745 ms (total bars in history 576660)
Time: 22.7 seconds

We see that the time of the first one (26.5 sec) is 15% worse than the second one (22.7 sec), though the PR of the first one is 70, and that of the second one is 49.

 

Alternatively:

Maybe one supports processor-specific instructions and the other does not.
And this is not taken into account in the calculation.
 

Intel Core2 Quad @2.40GHZ, 4095 MB, PR75

EURUSD,M1: 7143115 ticks (212232 bars) generated within 22449 ms (total bars in history 576660)

Z.Y. The number of ticks is different, by 15 different.

By the way, this may be the effect of the OS. I have Win 7 -64.

 
PR will be higher in the x32 terminal than in the x64 terminal.
 

Intel P4-3.06/512/533, 1024, PR54

EURUSD,M1: 7143132 ticks (212232 bars) generated within 42922 ms (total bars in history 576660)

 

strange. everyone has a different number of simulated ticks. but the number of bars is the same. i wonder if the reports are the same ?

Net profit: - 404.65 Total profit: 1,781.53 Total loss: -2,186.18

Files:
 
Prival:

strange. everyone has a different number of simulated ticks. but the number of bars is the same. i wonder if the reports are the same ?

Net profit: - 404.65 Total profit: 1,781.53 Total loss: -2,186.18

Yes, my results are the same to the penny, although the number of ticks is 15 more than yours.
 

Intel Core2 Duo E6550 @ 2.33Ghz, 2038 MB, PR92
EURUSD,M1: 7143130 ticks (212232 bars) generated within 19250 ms (total bars in history 576660)
WinXP

Profits match

 


EURUSD,M1: 7143132 ticks (212232 bars) generated within 27877 ms (total bars in history 576660)
Core2 Duo T5750 2.00 GHz, 2048 Mb PR54

W7 x64

MT5 x64

Matches

 
Batter has beaten everyone again ))) (memory is less, winds seem to be "worse", processor is a bit cooler, but not that much) and the time is minimal. I wonder if anyone has E8500 or E8600 please post the results
 

EURUSD,M1: 7143130 ticks (212232 bars) generated within 20296 ms (total bars in history 576660)
Core i3 M 330 2.13 GHz 2804 Mb PR58

W7 x64

MT5 x64

Reason: