The Daily Byte
Breaking
Welcome to The Daily Byte — real stories from across Pakistan.
Saturday, 21 November 2026 · Pakistan

The Daily Byte

Stories from Pakistan that deserve to be heard
Minor World

Nvidia, AMD and Google Back Open SGLang AI Inference

By The Daily Byte ·13 Nov 2026 ·2 min read
A long aisle in a modern data centre lined with black server racks lit by blue indicator lights

LMSYS Org has launched a new consortium bringing together eight of the biggest names in AI hardware and cloud computing to back SGLang, the open-source engine that runs large AI models in production. The founding members are Nvidia, AMD, Intel, Google, Meta, Cloudflare, Modal and RadixArk, Business Wire reported.

The SGLang Consortium was announced on 12 November in the opening keynote of the first SGLang Summit in San Francisco, which runs on 12 and 13 November at the Fort Mason Center for Arts and Culture. Members will work with SGLang's maintainers to shape the project's roadmap and turn shared priorities into joint engineering work.

What the group will work on

The consortium will focus on four areas: supporting frontier models, improving reasoning and multimodal workloads, scaling distributed inference, and optimising performance across hardware platforms. Each founding member names a technical representative and joins two roadmap meetings a year, while the maintainers keep control of technical decisions and code acceptance.

Why inference matters now

SGLang runs on hundreds of thousands of graphics processors and powers some of the largest production inference deployments in the world, for companies including Meta, Microsoft, Google, Amazon and Thinking Machines Lab, Daily AI Brief reported. LMSYS cites a Deloitte estimate that inference will account for about two-thirds of all AI computing in 2026, up from about half in 2025, which is why rival chipmakers and cloud firms are now cooperating on the software that runs models after training.