Computing

PyTorch 1.12 includes GPU acceleration on Apple Silicon.

I recently build a hyper converged node for home use. It virtualizes networking, storage and compute. To achieve that it uses Proxmox, ZFS, docker and pfSense.

This is a guide for increasing GPU training throughput given a batch of training run, as opposed to increasing instantaneous throughput for one training run. Obviously, doing the latter can increase training throughput, but that is not in the scope of this guide.