SGLang Enables Dual-Node Local Execution of Inkling-Small

SGLang now supports running Thinky Machines' Inkling-Small model locally across two NVIDIA DGX Spark systems connected via ConnectX-7, achieving an inference speed of 24 tokens per second.

2026-08-01 ~ 2026-08-03 · 2 related posts

• SGLang Supports Inkling-Small on Dual DGX Spark, Hits 24 tok/s — ying11231 · 2026-08-01

1 near-duplicate retellings: BanghuaZ