I've got an AMD FX-8350 CPU and a NVIDIA GeForce GTX 960 GPU on a PC with OpenSUSE Leap x86_64. I want to work with CUDA and OpenCL on it, so I first installed Cuda Toolkit 7.5. That worked fine.
Then, I installed AMD APP SDK 3.0, because I want to work OpenCL on the AMD CPU. That worked fine, it seems. $AMDAPPSDKROOT
echoes /opt/AMDAPPSDK-3.0
and $LD_LIBRARY_PATH
echoes /opt/AMDAPPSDK-3.0/lib/x86_64/
.
After that, I compiled and ran a basic deviceQuery code (basically this one, but modified to show platform name) with the following command:
gcc -L$AMDAPPSDKROOT/lib/x86_64/ -I$AMDAPPSDKROOT/include anotherDeviceQuery.c -lOpenCL
The code compiles and executes, but the output only shows the NVIDIA GPU info.
Platform NVIDIA CUDA found.
1. Device: GeForce GTX 960
1.1 Hardware version: OpenCL 1.2 CUDA
1.2 Software version: 352.39
1.3 OpenCL C version: OpenCL C 1.2
1.4 Parallel compute units: 8
I can't seem to find any solution. What could the problem be?