How to utilize the second GPU in dual GPU PC?
My application requires to work on Tensorflow 2.0+ with python 3.7 version. My Windows10 Desktop PC have dual GPU , NVIDIA NVS 310 and NVIDIA Tesla K40c . When I checked the driver for these , I found compatible version 392.59 Quadro for NVS 310 and version 412.36 for Tesla K40c. When driver version 392.59 is installed , for Tensorflow , only CUDA 8.0 recognizing the GPU’s availability but Tensorflow 2.0 is required CUDA 10.0+ version to install. In my Tensorflow python code, I mention “CUDA_ENABLE_DEVICES”=“1” , to enable Tensorflow CUDA for second GPU (ie:Tesla), but I found error:CUDA version is not compatible .
Now I installed driver version 412.36 , after installation , there is no such GPU in task manager display list and “NVIDIA Display setting are not available” .
Can anyone suggest how can I utilize second GPU for Tensorflow 2.0+ in this case. Thanks in advance.