---
cited_by:
  - '[[thoughts/papers/kimiteam2026kimik3openfrontier#13-mla-k3-variant-nvidiamlapy|kimik3 §1.3]]'
  - '[[thoughts/inference/kimi-k3-prs#4-pr-reading-order|kimik3 §4]]'
date: '2026-11-11'
kind: pr
local:
  - file:///Users/aarnphm/workspace/vllm/vllm/model_executor/layers/linear.py
merged: '2026-08-28'
number: 54015
source:
  - https://github.com/vllm-project/vllm/blob/6ec92bcbc8ef9af25cb4834ba3d922115792fbeb/vllm/models/kimi_k3/nvidia/mla.py
  - https://github.com/vllm-project/vllm/blob/6ec92bcbc8ef9af25cb4834ba3d922115792fbeb/vllm/model_executor/layers/linear.py
  - https://github.com/vllm-project/vllm/blob/6ec92bcbc8ef9af25cb4834ba3d922115792fbeb/vllm/models/kimi_k3/nvidia/model.py
  - https://github.com/vllm-project/vllm/blob/6ec92bcbc8ef9af25cb4834ba3d922115792fbeb/vllm/models/kimi_k3/nvidia/mtp.py
state: merged
tags:
  - vllm
  - vllm-pr
  - new-architectures
title: '#54015 [Kimi-K3] Merge MLA gate into QKV-A projection'
topic: new-architectures
url: https://github.com/vllm-project/vllm/pull/54015
created: '2026-11-11'
modified: '2026-11-11'
published: '2026-11-11'
pageLayout: default
slug: hinterland/papers/maps/prs/54015
permalink: https://aarnphm.xyz/hinterland/papers/maps/prs/54015.md
generator:
  quartz: v4.6.0
  hostedProvider: Cloudflare
  baseUrl: aarnphm.xyz
full: https://aarnphm.xyz/llms-full.txt
---
This PR merges MLA’s `g_proj` (used for output gating) into QKV_a projection to create a larger GEMM. We don’t expect this to have a large impact as the original QKV_a gemm is already quite large. The main purpose of this PR is to prepare… · full dump: [[thoughts/inference/vllm-prs/54015|#54015]]

