For deep learning I can't specifically comment, but for GPGPU I can: give AMD your money and use OpenCL and the upcoming Vulkan. It worked great for me, and gi

For deep learning I can't specifically comment, but for GPGPU I can: give AMD your money and use OpenCL and the upcoming Vulkan. It worked great for me, and given Nvidia deliberately de-rate their support for OpenCL to push CUDA, I feel it's important to signal with your money that given the choice between flexible and open standards vs. proprietary, I'll support openness.