Intel released llm-scaler-vllm 0.14.0-b8.1, a Docker-based deployment setup for running large language models on Intel Arc Graphics hardware via vLLM. The update adds support for several new Qwen models: Qwen3.5-27B, Qwen3.5-35B-A3B, Qwen3.5-122B-A10B (in FP8 and INT4 formats), and Qwen3-ASR-1.7B. The project builds on Intel's Project Battlematrix driver work.

1m read timeFrom phoronix.com
Post cover image

Sort: