comp.lang.idl-pvwave archive: archive » Re: IDL performance and FFTs (was: call external speed)

Home » Public Forums » archive » Re: IDL performance and FFTs (was: call external speed)

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

Re: IDL performance and FFTs (was: call external speed) [message #12882 is a reply to message #12875]

Wed, 16 September 1998 00:00

roy.hansen
Messages: 8
Registered: September 1998

Junior Member

In article <6to8cu$bgm$1@readme.uio.no>, steinhh@ulrik.uio.no (Stein Vidar Hagfors Haugan) wrote:
>
> In article <35ff8db4.0@d2o203.telia.com> roy.hansen@triad.no
> (Roy E. Hansen) writes:
>> We did a small comparison of the FFT performance in IDL 5.1.1
>> compared with the Matlab 5.2 version for a PII-400 with Win-NT,
>> and found that Matlab was approx 4 times faster. We also found
>> that the FFT in IDL 5.1.1 was faster than in IDL 5.1 on an other
>> PII-400 with Win95.
>
> I'm surprised the difference to Matlab 5.2 was so large.

Well, we have now done a sligthly more accurate test with the
following time results for the double precision FFT:

1D 2D
--------------------------------------
IDL 5.1.0 9.9 2.58
IDL 5.1.1 5.9 1.91
Matlab 5.2 3.6 1.0
--------------------------------------

This means that Matlab is 1.9 times faster for 2D FFTs and
1.6 times faster for 1D FFTs than IDL 5.1.1 for this specific case.

The most surprising is the difference between IDL 5.1.0 and
IDL 5.1.1 - a performance gain of 1.68 for 1D FFTs.

There were less than 5% difference between Win-NT and Win95.

The numbers in the table is the execution time of (without the declaration)
1D: x = dcomplexarr(4096) & for i=0,999 do fft(x,-1,/DOUBLE)
2D: x = dcomplexarr(256,256) & for i=0,9 do fft(x,-1,/DOUBLE)

Report message to a moderator

[Message index]

		Re: IDL performance and FFTs (was: call external speed) By: stevenj on Thu, 17 September 1998 00:00
		Re: IDL performance and FFTs (was: call external speed) By: menakkis on Thu, 17 September 1998 00:00
		Re: IDL performance and FFTs (was: call external speed) By: David Kastrup on Thu, 17 September 1998 00:00
		Re: IDL performance and FFTs (was: call external speed) By: Karl Krieger on Thu, 17 September 1998 00:00
		Re: IDL performance and FFTs (was: call external speed) By: steinhh on Thu, 17 September 1998 00:00
		Re: IDL performance and FFTs (was: call external speed) By: stevenj on Wed, 16 September 1998 00:00
		Re: IDL performance and FFTs (was: call external speed) By: roy.hansen on Wed, 16 September 1998 00:00
		Re: IDL performance and FFTs (was: call external speed) By: Liam Gumley on Wed, 16 September 1998 00:00
		Re: IDL performance and FFTs (was: call external speed) By: steinhh on Wed, 16 September 1998 00:00

Previous Topic:	trinagle plots in IDL
Next Topic:	Anyone have CLEAN (astronomical "deconvolution" routine)?

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sun Nov 30 06:00:14 PST 2025

Total time taken to generate the page: 0.01362 seconds