AI Infra Guide
Learn the AI Infra stack from scratch (updated regularly)
From low-level CUDA programming to distributed training and inference optimization
Knowledge base
Structured notes on the core AI Infra stack
AI Infra learning path AI Infra learning path
A structured AI Infra study path and knowledge map
AI Infra prerequisites AI Infra prerequisites
GPU architecture, computer architecture and other groundwork
CUDA programming and kernel optimization CUDA programming and kernel optimization
The CUDA programming model, writing kernels and tuning them
Distributed training Distributed training
Data parallelism, model parallelism and training at scale
Inference optimization Inference optimization
Model compression, quantization and inference engine tuning
Performance analysis Performance analysis
Finding bottlenecks, profiling tools and tuning strategy
Blog
Recent posts · View all
About Caomaolufei
Works on AI Infra at a large internet company and wrote AIInfraGuide. Focuses on model training, inference and kernel optimization.
AIInfraGuide article update log
A record of every content update to the AIInfraGuide knowledge base, so readers can track what changed
Tutorial roadmap
The update plan for four modules and 29 chapters: what is coming, how often, and where to get it.