🎯 Continuous Batching

The items that name it 4

• Computation and Data Movement for Inference
  article2026-09-21

  Lets requests join or leave a batch independently as they finish, freeing capacity for new work; works best when the scheduler matches slots to context length and service objective.

• The Roadmap to Mastering LLM Inference Optimization
  article2026-09-21
• datawhalechina/zero-to-sglang
  course2026-08-27
• patchy631/time-to-first-token
  repo2026-08-02