Hello, I’m posting because I’ve encountered an odd issue and am unsure of how to go about fixing it. I’ve written a C++ program that is using CUDA to perform an FFT and IFFT using the built in FFT.
The program itself was working fine until today when I upgraded the GPU, swapping a GTX 660 TI for a GTX 1070. After the switch I updated CUDA (7.5 to 8.0) and the driver, but for some reason the program seems to stop when I try to create the plan for FFT and/or IFFT.
When the program reaches this point it seems to simply stop running, there is no crash or error code, it will simply sit until I end the program. If I comment out the two commands the program will run through, though skipping the FFT and IFFT. Also, if I swap back in the 660 TI the program will run fine again.
I’m hoping that someone might give me some idea as to the cause and (hopefully) solution to this issue. I’m not particularly knowledgable about C++ or CUDA so without some error I’m pretty much lost