[Numpy-discussion] fourier with single precision

Nils Wagner nwagner@iam.uni-stuttgart...
Mon Aug 6 02:09:54 CDT 2007


Lars Friedrich wrote:
> Hello,
>
> thanks for your comments. If I got you right, I should look for a 
> FFT-code that uses SSE (what does this actually stand for?), which means 
> that it vectorizes 32bit-single-operations into larger chunks that make 
> efficient use of recent CPUs.
>
>   
http://en.wikipedia.org/wiki/Streaming_SIMD_Extensions
 
Nils



More information about the Numpy-discussion mailing list