AI/ ai · video generation · diffusion models · research

A Leaner Way to Generate Long Videos with Diffusion Models

Researchers swapped attention layers for state-space models in video diffusion, cutting memory costs without sacrificing quality on sequences up to 256 frames.

Generating long videos with diffusion models just got significantly cheaper — at least in terms of GPU memory.

A research team has published work replacing the attention layers typically used in video diffusion models with structured state-space models (SSMs), specifically a variant called Mamba. The core problem with attention is that its memory and compute costs grow quadratically with sequence length — fine for short clips, punishing for anything longer. SSMs scale linearly instead. The team also found that bidirectional SSMs, which process sequences in both directions rather than only forward, do a better job capturing temporal context in video data, echoing findings from image generation research.

For sequences of up to 256 frames, their SSM-based models use less memory to reach the same video quality score — measured by Frechet Video Distance — as comparable attention-based models, and often outperform them at equivalent memory budgets. That matters because memory is currently one of the harder ceilings on practical long-video generation, and any architecture that relaxes it opens the door to longer, higher-resolution outputs without proportionally larger hardware.

SSMs have been circling the transformer-replacement conversation for a while now, with most of the momentum in language modeling. Seeing the gains carry into video generation — a domain with far longer effective sequence lengths — is the more interesting signal here. Whether it holds at the resolutions and durations that commercial video tools actually target remains to be seen.

TR

The Revision

Written by an AI system from the public sources credited above. How we write →