Is there a discussion about the current FFT algorithm used by CVI 2012? Comparison to FFTW and Intel PPP would be helpful alsi if available.

[blakney]

One of my main application is pretty FFT intensive and therefore FFT performance is always on my mind.

 A few years back I recall a white paper (or similar) about CVI implementing the Intel MKL algorithm for improved performance. Is there a current white paper or similar for Intel processor based FFT under CVI 64-bit  for largish arrays (8 to 64 Gsample) ?

Also experiences with external FFT algorithms is appreciated. I have been using the now defunct Intel SPL for 32 bit apps (forever) and the Intel PPP for 64 bit apps (recently) but I am now having some deployment issues with the 64 - bit PPP packages. I have not determined if the PPP is to blame but it definitely seems to shut down execution profiling. ( I understand that the use of Intel PPP is not supported....)

[Jacob_R]

Hello blakney,

To my knowledge we do not have any white papers directed at Intel processor based FFTs under LabWindows/CVI 64-bit. Could you post a link to the whitepaper implementing the Intel MKL algorithm in LabWindows/CVI?