News: 0001645605

  ARM Give a man a fire and he's warm for a day, but set fire to him and he's warm for the rest of his life (Terry Pratchett, Jingo)

Intel-Scaler-vLLM 0.21.0-b1 Delivers Latest Features For vLLM On Intel GPUs

([Intel] 19 Minutes Ago Intel-Scaler-vLLM 0.21)


Released this morning was Intel's newest version of Intel-Scaler-vLLM as their Docker-based solution providing an optimized vLLM stack for execution on Intel Arc (Pro) graphics hardware.

Intel-Scaler-vLLM 0.21.0-b1 rebases against upstream vLLM 0.21 while also now experimentally supporting XPU graph with their software stack.

The new Intel-Scaler-vLLM release also adds support for diffusiongemma-26B-A4B-it, gemma-4-12B-it, gemma-4-31B-it, and gemma-4-26B-A4B-it.

This is a fairly straight-forward Intel Scaler vLLM update in simply refreshing against newer vLLM and adding support for additional Gemma 4 models for the bulk of the changes. If vLLM on Intel graphics interest you, learn more via [1]GitHub or try out the new release by heading on over to [2]Docker Hub .



[1] https://github.com/intel/llm-scaler/releases/tag/vllm-0.21.0-b1

[2] https://hub.docker.com/layers/intel/llm-scaler-vllm/0.21.0-b1/images/sha256-5d87be271e4db54539f1dbb29c071e9122f4e57b74594dbb26a55d27a569d780



A dwarf is passing out somewhere in Detroit!