Mac VMs get 11-16x faster AI speed with GPU trick

Original: Apple Silicon and macOS VMs: 11–16× Faster LLM Inference with Llama.cpp

Source: github.com ↗