Browse by topic

llm-compute

Explore the topics and references in this collection.

Modern AI Chips & Data Center Compute Architectures Cheat Sheet: CPU, GPU, TPU, LPU, NPU, DPU & Global Silicon Selection

A high-density reference guide to modern AI data center and edge compute silicon: covering CPU, GPU, TPU, LPU, NPU, DPU architectures, global silicon ecosystems (Huawei Ascend, Cambricon, Hygon, AWS Trainium, Meta MTIA, Microsoft Maia), Prefill/Decode disaggregation topologies, LLM VRAM/bandwidth sizing formulas, and nvidia-smi/npu-smi cluster operations.

Open cheat sheet