| title | TideGS: Scalable Training of Over One Billion 3D Gaussian Splatting Primitives via Out-of-Core Optimization |
|---|---|
| authors | Chonghao Zhong, Linfeng Shi, Hua Chen, Tiecheng Sun, Hao Zhao, Binhang Yuan, Chaojian Li |
| date | 2026-05-19 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2605.20150 |
| affiliation | Sponge Computing Lab at HKUST |
| upvotes | 6 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf code |
| abstract | Training 3D Gaussian Splatting (3DGS) at billion-primitive scale is fundamentally memory-bound: each Gaussian primitive carries a large attribute vector, and the aggregate parameter table quickly exceeds GPU capacity, limiting prior systems to tens of millions of Gaussians on commodity single-GPU hardware. We observe that 3DGS training is inherently sparse and trajectory-conditioned: each iteration activates only the Gaussians visible from the current camera batch, so GPU memory can serve as a working-set cache rather than a persistent parameter store. Building on this insight, we introduce TideGS, an out-of-core training framework that manages parameters across an SSD-CPU-GPU hierarchy via three synergistic techniques: block-virtualized geometry for SSD-aligned spatial locality, a hierarchical asynchronous pipeline to overlap I/O with computation, and trajectory-adaptive differential streaming that transfers only incremental working-set deltas between iterations. Experiments show that TideGS enables training with over one billion Gaussians on a single 24 GB GPU while achieving the best reconstruction quality among evaluated single-GPU baselines on large-scale scenes, scaling beyond prior out-of-core baselines (e.g., approximately 100M Gaussians) and standard in-memory training (e.g., approximately 11M Gaussians). |
cite
@misc{zhong2026tidegs,
title={TideGS: Scalable Training of Over One Billion 3D Gaussian Splatting Primitives via Out-of-Core Optimization},
author={Chonghao Zhong and Linfeng Shi and Hua Chen and Tiecheng Sun and Hao Zhao and Binhang Yuan and Chaojian Li},
year={2026},
eprint={2605.20150},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2605.20150}
}