# Chipmunk: Training-Free Acceleration of Diffusion Transformers with Dynamic Column-Sparse Deltas

arXiv: https://arxiv.org/abs/2506.03275 · HTML: https://mlsanity.com/paper/2506.03275/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Placed in Diffusion foundations and sampling, one of 14 directions in Diffusion models.
- Builds on 50 of its 60 references that are in the map.
- First posted on arXiv 3 Jun 2025.

## Placed in

- Diffusion models / [Diffusion foundations and sampling](https://mlsanity.com/diffusion/diffusion-foundations-and-sampling/)

## Builds on (most cited first, 30 of 50)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [Denoising Diffusion Implicit Models](https://mlsanity.com/paper/2010.02502/) (2020)
- [Deep Unsupervised Learning using Nonequilibrium Thermodynamics](https://mlsanity.com/paper/1503.03585/) (2015)
- [Scalable Diffusion Models with Transformers](https://mlsanity.com/paper/2212.09748/) (2022)
- [Scaling Rectified Flow Transformers for High-Resolution Image Synthesis](https://mlsanity.com/paper/2403.03206/) (2024)
- [Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flow](https://mlsanity.com/paper/2209.03003/) (2022)
- [FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness](https://mlsanity.com/paper/2205.14135/) (2022)
- [Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer](https://mlsanity.com/paper/1701.06538/) (2017)
- [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://mlsanity.com/paper/2101.03961/) (2021)
- [DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps](https://mlsanity.com/paper/2206.00927/) (2022)
- [Consistency Models](https://mlsanity.com/paper/2303.01469/) (2023)
- [Big Bird: Transformers for Longer Sequences](https://mlsanity.com/paper/2007.14062/) (2020)
- [Reformer: The Efficient Transformer](https://mlsanity.com/paper/2001.04451/) (2020)
- [HunyuanVideo: A Systematic Framework For Large Video Generative Models](https://mlsanity.com/paper/2412.03603/) (2024)
- [ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generation](https://mlsanity.com/paper/2304.05977/) (2023)
- [CogVideo: Large-scale Pretraining for Text-to-Video Generation via Transformers](https://mlsanity.com/paper/2205.15868/) (2022)
- [GenEval: An Object-Focused Framework for Evaluating Text-to-Image Alignment](https://mlsanity.com/paper/2310.11513/) (2023)
- [PixArt-α: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesis](https://mlsanity.com/paper/2310.00426/) (2023)
- [Open-Sora: Democratizing Efficient Video Production for All](https://mlsanity.com/paper/2412.20404/) (2024)
- [FlashAttention-3: Fast and Accurate Attention with Asynchrony and Low-precision](https://mlsanity.com/paper/2407.08608/) (2024)
- [Quest: Query-Aware Sparsity for Efficient Long-Context LLM Inference](https://mlsanity.com/paper/2406.10774/) (2024)
- [Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach](https://mlsanity.com/paper/2502.05171/) (2025)
- [DeepCache: Accelerating Diffusion Models for Free](https://mlsanity.com/paper/2312.00858/) (2023)
- [MAGI-1: Autoregressive Video Generation at Scale](https://mlsanity.com/paper/2505.13211/) (2025)
- [Timestep Embedding Tells: It's Time to Cache for Video Diffusion Model](https://mlsanity.com/paper/2411.19108/) (2024)
- [FlashInfer: Efficient and Customizable Attention Engine for LLM Inference Serving](https://mlsanity.com/paper/2501.01005/) (2025)
- [Deja Vu: Contextual Sparsity for Efficient LLMs at Inference Time](https://mlsanity.com/paper/2310.17157/) (2023)

## Built on by (most cited first, 8 of 8)

- [A Survey on Cache Methods in Diffusion Models: Toward Efficient Multi-Modal Generation](https://mlsanity.com/paper/2510.19755/) (2025)
- [HiCache: A Plug-in Scaled-Hermite Upgrade for Taylor-Style Cache-then-Forecast Diffusion Acceleration](https://mlsanity.com/paper/2508.16984/) (2025)
- [Universal Properties of Activation Sparsity in Modern Large Language Models](https://mlsanity.com/paper/2509.00454/) (2025)
- [AdaCluster: Adaptive Query-Key Clustering for Sparse Attention in Video Generation](https://mlsanity.com/paper/2604.18348/) (2026)
- [SparVAR: Exploring Sparsity in Visual AutoRegressive Modeling for Training-Free Acceleration](https://mlsanity.com/paper/2602.04361/) (2026)
- [DC-DiT: Adaptive Compute and Elastic Inference for Visual Generation via Dynamic Chunking](https://mlsanity.com/paper/2603.06351/) (2026)
- [LinCa: Accelerating Diffusion Models via Learnable Decomposed Feature Caching](https://mlsanity.com/paper/2608.17973/) (2026)
- [Hardware Acceleration of Block-Diffusion LLM for Edge Devices](https://mlsanity.com/paper/2609.01084/) (2026)

## Abstract (arXiv)

Diffusion Transformers (DiTs) have achieved state-of-the-art performance in high-quality image and video generation but incur substantial compute cost at inference. A common observation is that DiT latent noise vectors change slowly across inference steps, which suggests that the DiT compute may be redundant across steps. In this paper, we aim to speed up inference by reducing this redundancy, without additional training. We first study how activations change between steps in two state-of-the-art open-source DiTs. We find that just 5-25% of the values in attention and MLP explain 70-90% of the change in activations across steps. This finding motivates our approach, Chipmunk, which uses dynamic sparsity at inference time to recompute only the fastest-changing intermediate activations, while caching the rest. Dynamic sparsity introduces two systems challenges: (1) sparse attention and MLP operations tend to underutilize GPU tensor cores; and (2) computing dynamic sparsity patterns at runtime and caching activations both introduce overhead. To address these challenges, Chipmunk first uses a voxel-based reordering of input tokens to introduce column-wise sparsity. We implement column-sparse kernels utilizing efficient sparse gathers from global to shared GPU memory, achieving a 9.3x speedup at 93% sparsity compared to highly-optimized dense baselines. Second, Chipmunk overlaps the computation of sparsity patterns and cache updates with other parts of the computation (e.g., second layer of the MLP) to hide the extra latency. Chipmunk achieves up to 2.16x speedup on HunyuanVideo and 1.41x on FLUX.1-dev without compromising generation quality. Furthermore, we show that Chipmunk can be stacked on top of full step caching, achieving a 3.72x speedup on HunyuanVideo, a 2.67x speedup on WAN2.1, and a 2.25x speedup on FLUX.1-dev with minimal quality impact.