---
cited_by:
  - '[[thoughts/papers/qiu2026designqwen38nextarchitectureevaluation#11-qsa--qwen-sparse-attention|qwen4next §1.1]]'
  - '[[thoughts/inference/qwen3.8-serving#4-pr-reading-order-vllm-main-git-show-sha|qwen4next §4]]'
date: '2026-11-11'
kind: pr
local:
  - file:///Users/aarnphm/workspace/vllm/.buildkite/test_areas/models_basic.yaml
merged: '2026-10-01'
number: 57097
source:
  - https://github.com/vllm-project/vllm/blob/c9578f10ab3eaf252742960db3fd08f11cf6d25a/vllm/models/qwen4_exp/nvidia/ops/qsa_prepare.py
  - https://github.com/vllm-project/vllm/blob/c9578f10ab3eaf252742960db3fd08f11cf6d25a/vllm/models/qwen4_exp/nvidia/qsa.py
  - https://github.com/vllm-project/vllm/blob/c9578f10ab3eaf252742960db3fd08f11cf6d25a/vllm/models/qwen4_exp/nvidia/indexer_qsa.py
  - https://github.com/vllm-project/vllm/blob/c9578f10ab3eaf252742960db3fd08f11cf6d25a/.buildkite/test_areas/models_basic.yaml
state: merged
tags:
  - vllm
  - vllm-pr
  - new-architectures
title: '#57097 [Qwen3.8-Flash-Next] Fuse main QK-norm/RoPE/gate and KV-cache write into the QSA pre-indexer launch'
topic: new-architectures
url: https://github.com/vllm-project/vllm/pull/57097
created: '2026-11-11'
modified: '2026-11-11'
published: '2026-11-11'
pageLayout: default
slug: hinterland/papers/maps/prs/57097
permalink: https://aarnphm.xyz/hinterland/papers/maps/prs/57097.md
generator:
  quartz: v4.6.0
  hostedProvider: Cloudflare
  baseUrl: aarnphm.xyz
full: https://aarnphm.xyz/llms-full.txt
---
1. `fused_qk_rmsnorm_rope_gate`: main-attention Q/K RMSNorm, RoPE and the gate split. · full dump: [[thoughts/inference/vllm-prs/57097|#57097]]

