First SGLang Summit draws 2,000 devs, launches Consortium with 9 members and SGLang v1.0
ying11231 · x · 2026-11-18
The first SGLang Summit (Nov 12–13, Fort Mason, San Francisco) gathered 2,000+ developers, researchers, and industry leaders. Key announcements:
- SGLang Consortium launched with nine founding members to build open AI inference infrastructure together.
- SGLang v1.0 released, setting a new milestone for production inference.
- Miles v0.2 expands reinforcement learning across hardware and workloads.
- Hardware reach widened: NVIDIA NIM Certified microservices, Intel collaboration, and an AWS Trainium preview.
- A call for security partners.
Talks and breakouts covered serving frontier models, production lessons, RL at scale, and designing the inference engine for the agentic era, plus hands-on workshops and a GPU expo.
More from Infra
- SGLang hits v1.0 as NVIDIA, AMD, Intel, Google and Meta form its consortium — nikola_mr64990 · 2026-11-18
- Report: Nvidia sales staff introduced Alice Huang to Alibaba pre-ban, pre-ChatGPT — HanchungLee · 2026-11-18
- Nvidia's CFO walks back defense of 'circular financing' investments — rohanpaul_ai · 2026-11-18
- DeepSeek V4.1 Flash 1017 now available for post-training on Together AI with custom RL/SFT loops — togethercompute · 2026-11-18
- Editing a 10-min video with AI costs 20-30M tokens, 20-min barely costs more — MatthewBerman · 2026-11-18
- Nvidia, with ~$100B in AI stakes, will require solid offtake before funding buyers — rohanpaul_ai · 2026-11-18