Version b11702, released Oct 17 2026, includes a new MxFP4 MoE bin kernel that speeds up LLM inference on Adreno X1 devices.
NoteBy OossaPublished by Oossa: 1 min read
The ggml‑org/llama.cpp project released version b11702 on Oct 17 2026. The update adds an OpenCL kernel called `gemm_moe_mxfp4_q8_1_dp4a_ila_ldck` for Qualcomm Adreno X1 GPUs. The change is part of a set of OpenCL tweaks, including whitespace fixes and DP4A support for X1E. Binary builds for macOS, iOS, and several Ubuntu flavors are also provided.
The kernel lets developers run LLMs faster on Android phones that use Adreno X1 GPUs.
Models
Beijing's new draft limits exports of large AI models and 23 AI tools, but fast‑tracks open‑source releases.
Models
Gemini 4 Argon scored 97.5% on the toughest general‑problem‑solving test, second only to OpenAI's GPT‑6 Astra.
Models
The Linux kernel maintainer used an LLM to fix a cache‑contention bug, showing a 23‑times speed boost in synthetic tests.
Oossa · Newsletter
The week in AI, explained
Every Monday: the stories worth knowing, in plain language. Free, no spam.
