On: Using TensorFlow in Windows with a GPU

[eva1984]

Will you are paying a premium to Nvidia not only for the hardware, but more for their CUDA libraries, or in fancier term, ecosystem.

Most DL frameworks are banking on cudnn, which I didn't see AMD or FPGA's equivalent come anytime soon.

[westoncb → eva1984]

I wonder if OpenGL's compute shaders (https://www.khronos.org/opengl/wiki/Compute_Shader) would be comparable to CUDA for this sort of thing...

[iaw → eva1984]

That is one thing I appreciate, everything from Nvidia's side works perfectly out of the box. I've only ever had undocumented issues on the FOSS side.