SGLang Summit 2026 agenda: Meta, Google, NVIDIA reveal production-scale SGLang playbooks
ying11231 · x · 2026-11-11
SGLang Summit 2026 (Nov 12-13, Fort Mason, SF) features Meta's Lianmin Zheng on serving billions via SGLang, LinkedIn running ads-ranking models at 10K+ QPS with <80ms latency, NVIDIA porting SGLang to Vera Rubin, Google deploying SGLang-TorchTPU, Alibaba extending HiCache for long-running agents, Perplexity's minimalist trillion-parameter trainer, and sessions from Cloudflare, Modal, Decagon, HeyGen and MiniMax.
Related event: First SGLang Summit Set for San Francisco with Meta, NVIDIA and Alibaba(3 posts)→
More from Infra
- Kevin Bass shows off 600GB-RAM homelab cluster for running coding agents — kevinnbass · 2026-11-11
- Token factory and data vendor revenue acceleration validates the 'token stream' investment thesis — Ronangmi · 2026-11-11
- Modal engineer: inference consensus converging on disagg + wide EP + KV storage — dejavucoder · 2026-11-11
- Tesla installs 133 MW / 266 MWh Megapack at Giga Texas, up to 30 units a day — XFreeze · 2026-11-11
- Inside the next-gen data center cooling tech reusing 750 gallons of water per minute — TinfoilTricorn · 2026-11-11
- Darpinian: Tesla should wait for the AI5 chip before scaling Cybercab production — Darpinian · 2026-11-11