FFT double precision Does anybody use it ?


Do you know if CuFFT library is compliant with double precision ? Any performance idea ??

I’ve heard that a new version of CUFFT will be released with better performances. Does anybody know when ?

Thanks for your answers, you’re doing a great job here :)


It looks to me like the CuFFT library does not support double precision. However, I would be very interested in hearing whether there are any plans to produce such a version, which would only be useful with ( compute capability > 1.3 ) devices.