Week 44, 2026
- Notes about Cell-Based Architecture
- OCP 2026 - Co-Designing an AI Cluster - How power, cooling, and networking shape AI infrastructure, and the tradeoffs between standardization and tight integration.
- Enabling Efficient Serverless Inference Serving for LLM (Large Language Model) in the Cloud - A review of cold starts in serverless LLM inference, including ServerlessLLM’s checkpoint loading, live migration, and scheduling techniques.