GPU-accelerated code on CPU-only environments -Remote GPU Kernel Execution
1–2 of 2 posts
Re: GPU-accelerated code on CPU-only environments -Remote GPU Kernel Execution
#2We built this feature in our GPU Hypervisor, which cleanly separates your user-space ML environment from the GPU runtime, so you can code locally, run remotely, and execute kernels remotely on GPU hosts. Would love to hear how this impacts your ML platforms. Pls comment or DM me.