# Intel releases LLM-Scaler-vLLM PV 1.5

Digest AI · Enterprise & Industry · published 2026-10-16T20:16:18Z

Canonical: https://digestai.news/story/intel-releases-llm-scaler-vllm-pv-1-5

## Summary

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.

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.

## 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.

## Why it matters

Intel’s update brings higher‑level vLLM and PyTorch support to Arc GPUs, expanding model compatibility and improving inference speed. The new quantization options and faster token times help developers deploy large‑language models more efficiently on Intel hardware, potentially lowering costs and expanding AI adoption in edge and cloud environments.

## Sources

1. [Intel LLM-Scaler-vLLM PV 1.5 brings big improvements](https://phoronix.com/news/Intel-LLM-Scaler-vLLM-PV-1.5) (phoronix.com, 2026-10-16)

## Cite

Digest AI, "Intel releases LLM-Scaler-vLLM PV 1.5", 16 October 2026, https://digestai.news/story/intel-releases-llm-scaler-vllm-pv-1-5

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/intel-releases-llm-scaler-vllm-pv-1-5.json