Skip to content

SIG-LLM-Infra ​

Objective ​

Improve scale, reliability, and cost-efficiency of LLM infrastructure.

Key Topics ​

  • Network topology and communication optimization
  • Engineering practices for large-scale parallelism
  • Cost and reliability governance for inference services

Current Status ​

  • Chair: open call
  • Kick-off meeting: in preparation
  • Draft: in preparation — community review welcome
  • GitHub: to be announced

AI-HPC Expert · Contact: openaihpc@gmail.com