I have a project which I’m keen to use CUDA for FIR Signal Processing. I have mocked the Kernels up in a CUDA project which is fine. But I’m forced to use another framework for the main application.
I have added the correct library linking and headers and can successfully make calls such as cudaSetDevice(0) so I know that’s all working. I have also added
To my project file which exposes the extra CUDA functionality within my project. Everything compiles
When I add my kernel though I get that C2059 syntax error ‘<’ - which I know is symptomatic of it using NVCC. I have checked my FIRKernal.cu file Item type is set to CUDA / C++ but I’m not sure what else to do.
It’s at the point were my kernel is executed processFIR<<<1, size>>>(outSamples, inSamples, coefs, numSamples);
Any help appreciated