V — The Systems That Run Them → Chapter 23
FROM SYSTEMS TO FRONTIER ML

Runtimes & frameworks

PyTorch (eager vs compiled), CUDA, ONNX, MLX. What a 'kernel' is, the dispatch stack. A real CUDA dot product alongside its CPU SIMD twin.

§1 PyTorch eager vs compiled — the dispatch stack §2 CUDA, Triton, MLX — what a 'kernel' is §3 ONNX, GGUF, safetensors — model interchange formats

← ALL CHAPTERS