---
cited_by:
  - '[[thoughts/inference/deepseek-v4.1-serving#4-pr-reading-order|dsv41 §4]]'
date: '2026-11-11'
kind: pr
local: []
merged: '2026-09-16'
number: 56935
source:
  - https://github.com/vllm-project/vllm/blob/d6a1677d5504244c566eb900ca605cb5511f4ab4/vllm/models/deepseek_v41/nvidia/flash_mla_mega_attn.py
  - https://github.com/vllm-project/vllm/blob/d6a1677d5504244c566eb900ca605cb5511f4ab4/csrc/libtorch_stable/fused_deepseek_v4_qnorm_rope_kv_insert_kernel.cu
  - https://github.com/vllm-project/vllm/blob/d6a1677d5504244c566eb900ca605cb5511f4ab4/vllm/models/deepseek_v41/attention.py
  - https://github.com/vllm-project/vllm/blob/d6a1677d5504244c566eb900ca605cb5511f4ab4/vllm/models/deepseek_v41/common/ops/cache_utils.py
  - https://github.com/vllm-project/vllm/blob/d6a1677d5504244c566eb900ca605cb5511f4ab4/vllm/models/deepseek_v41/common/ops/fused_layout.py
  - https://github.com/vllm-project/vllm/blob/d6a1677d5504244c566eb900ca605cb5511f4ab4/vllm/models/deepseek_v41/common/ops/fused_compress_quant_cache.py
  - https://github.com/vllm-project/vllm/blob/d6a1677d5504244c566eb900ca605cb5511f4ab4/vllm/models/deepseek_v41/sparse_mla.py
  - https://github.com/vllm-project/vllm/blob/d6a1677d5504244c566eb900ca605cb5511f4ab4/vllm/models/deepseek_v41/nvidia/model.py
state: merged
tags:
  - vllm
  - vllm-pr
  - new-architectures
title: '#56935 [Model][DSv4.1] FlashMLA mega attention and the NVFP4 compressed KV cache'
topic: new-architectures
url: https://github.com/vllm-project/vllm/pull/56935
created: '2026-11-11'
modified: '2026-11-11'
published: '2026-11-11'
pageLayout: default
slug: hinterland/papers/maps/prs/56935
permalink: https://aarnphm.xyz/hinterland/papers/maps/prs/56935.md
generator:
  quartz: v4.6.0
  hostedProvider: Cloudflare
  baseUrl: aarnphm.xyz
full: https://aarnphm.xyz/llms-full.txt
---
Supersedes [#56344](https://github.com/aarnphm/aarnphm.github.io/issues/56344), which closed when its base branch `dsv41-optimized` was deleted. · full dump: [[thoughts/inference/vllm-prs/56935|#56935]]

