> I let the machine cool and ran a benchmark, then I ran another
> benchmark directly after the processor had been working hot
> for a while, then I ran another benchmark while the processor was
> hot and prime95.exe was set to Win2k's Real Time priority. The
> results:
And here's the results of David Meyer's three tests in another form:
: Best Time For Each FFT Length
Length :-----------------------------------------------------
: First Test Second Test^* Third Test
-------------:-----------------------------------------------------
256K FFT : 16.467 ms. 16.427 ms. 16.438 ms.
320K FFT : 21.976 ms. 22.041 ms. 21.964 ms.
384K FFT : 26.791 ms. 26.730 ms. 26.779 ms.
448K FFT : 32.243 ms. 32.180 ms. 32.179 ms.
512K FFT : 35.361 ms. 35.208 ms. 35.288 ms.
640K FFT : 47.979 ms. 47.758 ms. 47.877 ms.
768K FFT : 58.182 ms. 57.267 ms. 56.832 ms.
892K FFT : 71.367 ms. 71.628 ms. 71.646 ms.
1024K FFT : 78.093 ms. 76.722 ms. 76.920 ms.
1280K FFT : 111.840 ms. 112.063 ms. 111.058 ms.
1536K FFT : 141.851 ms. 140.377 ms. 140.468 ms.
1792K FFT : 181.142 ms. 182.122 ms. 180.737 ms.
> ^* it should be noted that the results of second test
> above may be slightly inflated due to the presense
> of temperature monitoring software
--
Lampros Kallenos
Lefkosia, Kypros
[EMAIL PROTECTED]
.
_________________________________________________________________________
Unsubscribe & list info -- http://www.ndatech.com/mersenne/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
