Add a concise note on how to think about GPUs

[Copilot]

This adds a dedicated note summarizing the core mental model from the JAX Scaling Book’s “How to Think About GPUs” section. The goal is to make the GPU architecture and performance intuition easier to scan from the notes site.

• New note

  • Added docs/misc/gpu.md with a concise overview of:
    • GPU as many SMs attached to HBM
    • Tensor Core / CUDA Core / SMEM roles
    • parallelism, Tensor Core utilization, memory hierarchy, and warp divergence
    • a practical GPU vs TPU comparison
    • a short performance-oriented takeaway for training/inference work

• Site navigation

  • Added the new page to Misc in mkdocs.yml so it is discoverable from the rendered notes site

• Content shape

  • Kept the page in the existing notes style: short, high-signal, source-linked summary instead of a long-form translation

# How to think about GPUs

**把 GPU 想成“一大块高带宽内存（HBM）+ 很多可以并行工作的计算单元（SM）”。**

> GPU 不是单个“超级核心”，而是很多相对独立的小核心，在一起吞吐大规模并行任务。

Warning

Firewall rules blocked me from connecting to one or more addresses (expand for details)

I tried to connect to the following addresses, but was blocked by firewall rules:

• jax-ml.github.io
  • Triggering command: /home/REDACTED/work/_temp/ghcca-node/node/bin/node /home/REDACTED/work/_temp/ghcca-node/node/bin/node --enable-source-maps /home/REDACTED/work/_temp/copilot-developer-action-main/dist/index.js (dns block)

If you need me to access, download, or install something from one of these locations, you can either:

• Configure Actions setup steps to set up my environment, which run before the firewall is enabled
• Add the appropriate URLs or hosts to the custom allowlist in this repository's Copilot coding agent settings (admins only)