FVAttn: Adaptive Sparse Attention with Runtime Load Balancing for Video Generation

• URL: http://arxiv.org/abs/2607.16190v1
• Date: Fri, 17 Jul 2026 17:59:32 GMT
• Title: FVAttn: Adaptive Sparse Attention with Runtime Load Balancing for Video Generation
• Abstract summary: Training-free sparse attention improves distributed execution efficiency of adaptive sparse attention under multi- GPU sequences.<n>Method reduces average load imbalance from 1.34 to 1.08 and delivers a $4.41times$ attention speedup over FlashAttention.<n>On step-distilled Wan2.2 I2V, method reduces average load imbalance from 1.34 to 1.08 and delivers a $4.41times$ attention speedup over FlashAttention.
• Score: 19.638077533543754
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Video Diffusion Transformers process long spatio-temporal sequences, making self-attention the main bottleneck in high-resolution video generation. Training-free sparse attention reduces this cost, but adaptive Top-$p$ routing creates uneven per-head workloads under multi-GPU sequence parallelism. The resulting workload heterogeneity turns sparse attention into a rank-level straggler problem. We present \method{}, a training-free sparse-attention system that improves the distributed execution efficiency of adaptive sparse attention under multi-GPU sequence parallelism. \method{} uses Top-$p$ routing, a Top-$k$ safety floor, and video-aware block organization as the sparse-routing frontend, then repairs the materialized mask at runtime. Runtime Load Balancing migrates a small number of heavy heads via P2P communication to shorten the current critical path. Slack-Aware Sparse Augmentation fills residual non-critical-rank slack with additional high-value blocks, while overlap hides scheduling and migration overhead behind existing computation. On step-distilled Wan2.2 I2V, \method{} reduces average load imbalance from 1.34 to 1.08 and delivers a $4.41\times$ attention speedup over FlashAttention, while achieving a $2.02$--$2.11\times$ DiT inference speedup with competitive video quality.

Related papers

• Fast Enough to Act: Spatio-Temporal Visual Token Merging for Low-Latency Robotic VLMs and VLAs [21.02858354959528]
  Vision-language models and vision-language action models endow the robot with unprecedented capabilities.<n>The input of video and high-resolution images yields a massive number of visual tokens, leading to extremely high inference latency and severely hindering the robot's real-time control.<n>We propose ST-Merge, a plug-and-play framework that efficiently fuses redundant tokens directly during the visual encoding phase.
  arXiv  Detail & Related papers  (2026-06-28T11:42:55Z)
• PARE: Pruning and Adaptive Routing for Efficient Video Generation [71.54959622788608]
  Video Diffusion Transformers (DiTs) generate high-quality videos but demand substantial compute due to wide blocks, deep architectures, and iterative sampling.<n>Recent methods reduce cost by compressing width, depth, or sampling steps, but typically commit to a fixed architecture that cannot adapt to individual inputs or denoising stages.<n>We propose PARE, which jointly compresses width and depth with structure-aware pruning and input-adaptive routing.
  arXiv  Detail & Related papers  (2026-05-26T17:43:25Z)
• HASTE: Training-Free Video Diffusion Acceleration via Head-Wise Adaptive Sparse Attention [68.95533683996236]
  Training-free sparse attention is attractive because it accelerates pretrained models without retraining.<n>Existing online top-$p$ sparse attention still spends non-negligible cost on mask prediction.<n>We show that these two overlooked factors limit the practical speed-quality trade-off of training-free sparse attention in Video DiTs.
  arXiv  Detail & Related papers  (2026-05-14T07:57:55Z)
• Expert Streaming: Accelerating Low-Batch MoE Inference via Multi-chiplet Architecture and Dynamic Expert Trajectory Scheduling [29.998645673111536]
  Mixture-of-Experts is a promising approach for edge AI with low-batch inference.<n>MoE sparsity and dynamic gating shift distributed strategies toward much finer runtime granularity.<n>Fully Sharded Expert Data Parallelism is a parallelization paradigm specifically architected for low-batch MoE computation.
  arXiv  Detail & Related papers  (2026-03-29T10:44:55Z)
• Spava: Accelerating Long-Video Understanding via Sequence-Parallelism-aware Approximate Attention [63.69228529380251]
  Spava is a sequence-parallel framework with optimized attention for long-video inference.<n>Spava delivers speedups of 12.72x, 1.70x, and 1.18x over FlashAttn, ZigZagRing, and APB, without notable performance loss.
  arXiv  Detail & Related papers  (2026-01-29T09:23:13Z)
• StreamDiffusionV2: A Streaming System for Dynamic and Interactive Video Generation [65.90400162290057]
  Generative models are reshaping the live-streaming industry by redefining how content is created, styled, and delivered.<n>Recent advances in video diffusion have markedly improved temporal consistency and sampling efficiency for offline generation.<n>Live online streaming operates under strict service-level objectives (SLOs): time-to-first-frame must be minimal, and every frame must meet a per-frame deadline with low jitter.
  arXiv  Detail & Related papers  (2025-11-10T18:51:28Z)
• SLA: Beyond Sparsity in Diffusion Transformers via Fine-Tunable Sparse-Linear Attention [88.47701139980636]
  In Diffusion Transformer (DiT) models, particularly for video generation, attention latency is a major bottleneck.<n>We find that attention weights can be separated into two parts: a small fraction of large weights with high rank and the remaining weights with very low rank.<n>We propose SLA, a trainable attention method that fuses sparse and linear attention to accelerate diffusion models.
  arXiv  Detail & Related papers  (2025-09-28T17:58:59Z)
• Astraea: A Token-wise Acceleration Framework for Video Diffusion Transformers [29.130090574300635]
  Video diffusion transformers (vDiTs) have made tremendous progress in text-to-video generation, but their compute demands pose a major challenge for practical deployment.<n>We introduce Astraea, a framework that searches for near-optimal configurations for vDiT-based video generation under a performance target.
  arXiv  Detail & Related papers  (2025-06-05T14:41:38Z)
• VORTA: Efficient Video Diffusion via Routing Sparse Attention [54.84294780326206]
  VORTA is an acceleration framework with two novel components.<n>It achieves an end-to-end speedup $1.76times$ without loss of quality on VBench.<n>It can seamlessly integrate with various other acceleration methods, such as model caching and step distillation, reaching up to speedup $14.41times$ with negligible performance degradation.
  arXiv  Detail & Related papers  (2025-05-24T17:46:47Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.