Open-weight & owned stack · Quantization efficiency

llama.cpp Adds IQ3S Vulkan Kernels

Sep 18, 2026, first seen via Training & inference tooling releases

llama.cpp releases b11035 through b11037 added IQ3_S MMQ Vulkan kernels and related two-byte shared-memory loading changes.

Read the original at github.comOpens the publisher's site in a new tab

Also covering this

b11037github.com, Sep 18 b11036github.com, Sep 18 b11034github.com, Sep 18 b11030github.com, Sep 18 b11028github.com, Sep 17 b11027github.com, Sep 17

More in Open-weight & owned stack

Xiaomi Publishes MiMo-V2.6 Reinforcement Learning DashboardSep 17 Gemma Models Surpass One Billion DownloadsSep 16 llama.cpp Fixes Remote-Code-Execution VulnerabilitySep 16 Alibaba Releases Occamy Open-Weight AI AgentSep 15 Apple Enables Local Foundation Models On MacOSSep 15