comp.lang.idl-pvwave archive: archive » Help, no improvement in FFT speed on a multiprocessor system

Home » Public Forums » archive » Help, no improvement in FFT speed on a multiprocessor system

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

Help, no improvement in FFT speed on a multiprocessor system [message #67916]

Sat, 05 September 2009 22:06

Marco
Messages: 5
Registered: April 2009

Junior Member

I'm running IDL 7.1 on a Linux 2.6. This is an HP quad processor with each
processors having 6 cores for 24 cores total.

Doing large 2-D FFTs (>8Kx8K) I get no benefit from the extra processors.
I can vary the number used in IDL from 24 down to 1, and see that the
number actually of processors actually showing a load is the correct number,
but from 4 to 24 threads, the speed is the same and no faster than Matlab,
which uses only a single processor out of the 24.

I've tried varied IDL_CPU_TPOOL_NTHREADS and IDL_CPU_TPOOL_MIN_ELTS, but
have not been able to improve the results.

Any suggestions?

Thanks in advance,

--
Marco
UCO Lick Observatory
Laboratory for Adaptive Optics

Report message to a moderator

Previous Topic:	IDL startup error message on Mac OS X Leopard
Next Topic:	Speedy Julia Set Fractals

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Wed Oct 08 15:47:54 PDT 2025

Total time taken to generate the page: 0.00462 seconds