Intel releases LLM-Scaler-vLLM PV 1.5
Intel software engineers released LLM-Scaler-vLLM PV 1.5, a Docker stack that runs large‑language models on Intel Arc Graphics. It also adds support for several Qwen and Gemma models, including Qwen3.6‑27B, Qwen3.8‑27B, Qwen3.6‑35B‑A3B, gemma‑4‑26B‑A4B‑it and gemma‑4‑31B‑it.
Key points
- Intel upgrades vLLM from 0.14 to 0.26 in LLM-Scaler-vLLM PV 1.5.
- New Qwen and Gemma models added, plus FP8/INT4 quantization and faster token times.
New quantization options—online FP8 and symmetric INT4—are now available, along with FP8 KV cache support. The stack delivers faster time‑to‑first‑token for Qwen and Gemma 4 models, improved MTP decoding, and better LoRA serving performance. Bug fixes round out the release, which can be downloaded from the project’s GitHub repository. The work is part of Intel’s Project Battlematrix and precedes the upcoming Crescent Island Intel GPU D Series.
Intel LLM-Scaler-vLLM PV 1.5 brings big improvements
phoronix.com · 16 October 2026Loading the full article…
This text was published by phoronix.com. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
1source- Reddit discussionreddit.com
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Enterprise & Industry
All →- HP integrates NVIDIA Open Agent Safety Platform into AI infrastructure · 1 src
- PostgreSQL 19 delayed to November as AI code reviews uncover extra bugs · 1 src
- Author argues AI fleets need organizational proprioception · 1 src
- Hogan Lovells Cadwalader builds custom Harvey AI agents for legal teams · 1 src
- Unite.AI warns MCP server security depends on upstream system flaws · 1 src
Comments
via GitHub Discussions