Bug?

2010-03-01, 04:43

(Vista64/i7965/prime95,25.11)...Hi,i wonder if anyone knows why some workers do more or less tests than others when using the same fft length?

thanks in advance:)

Uncwilly · 2010-03-01, 13:31

The exponent size also has an effect on how long it will take to complete a test. And often a single worker will be the one to suffer when the computer is doing other things.

2010-03-01, 17:46

Ok,thank you uncwilly!

CADavis · 2010-03-01, 20:24

yup other processes and the operating system take up different amount of each core so they are not equal in throughput. on my quad usually worker #2 (core 1) is fastest and worker #1 or #4 (cores 0 or 3) is slowest, but it's not always consistent.

2010-03-01, 17:46	#3
Unregistered 3²×947 Posts	goncourt of Ok,thank you uncwilly!

2010-03-01, 04:43	#1
Unregistered 11·13·29 Posts	Bug? (Vista64/i7965/prime95,25.11)...Hi,i wonder if anyone knows why some workers do more or less tests than others when using the same fft length? thanks in advance:)

2010-03-01, 13:31	#2
Uncwilly 6809 > 6502 """"""""""""""""""" Aug 2003 101×103 Posts 9844₁₀ Posts	The exponent size also has an effect on how long it will take to complete a test. And often a single worker will be the one to suffer when the computer is doing other things.

2010-03-01, 20:24	#4
CADavis Jul 2005 Des Moines, Iowa, USA 2×5×17 Posts	yup other processes and the operating system take up different amount of each core so they are not equal in throughput. on my quad usually worker #2 (core 1) is fastest and worker #1 or #4 (cores 0 or 3) is slowest, but it's not always consistent.