KernelWiki
Blackwell and Hopper GPU kernel optimization knowledge base
All Skills in This Repository (1)
Pure Emerald Level IndicatorsFrequently Asked Questions
FAQPage SchemaHow to install KernelWiki?โผ
Run `npx skills add mit-han-lab/KernelWiki --all -g -y` in your terminal to install the skill globally.
What is KernelWiki used for?โผ
It is a structured knowledge base for optimizing NVIDIA Blackwell (SM100) and Hopper (SM90) GPU kernels, covering hardware features, techniques, and real kernel case studies like FlashAttention-4 and DeepGEMM.
How do I search the KernelWiki knowledge base?โผ
Use the bundled scripts, for example `python3 scripts/query.py "warp specialization"` or `python3 scripts/get_page.py kernel-flash-attention-4`, to search by keyword, tag, architecture, or page ID.
Does KernelWiki include real code examples?โผ
Yes. Every technique, kernel, and language page is validator-enforced to include at least a compilable code snippet, and performance claims cite their original source.
Can I use KernelWiki without GPU programming experience?โผ
Yes. You can ask questions in plain English and your agent will query the wiki, follow cited sources, and return synthesized answers with confidence levels.
Related Repositories in Software Engineering
View All in Software Engineeringโopenclaw
Run a personal AI assistant across your devices and chat apps
superpowers
Gives coding agents a disciplined workflow from idea to merged code
react
AI agent skills for building, testing, and porting React core