I’m wondering how many possible reasons might lead to this error, because it’s really driving me crazy.
I’m trying to do some small 2D real-to-complex transformation on my 8800GTS. With same plan parameter and same input data, FFTW baseline program gives me the correct result but CUFFT keeps throwing out that error msg. The transformation size is 168168, therefore input size is 168168 and output is 170*168.
Anyone can give me some clue where the bug might be hiding? 1000000 thanks!