A Scalable PyTorch Abstraction for Multi-GPU Gaussian Splatting
Quick Answer
This paper introduces a multi-GPU Gaussian splatting method using a PyTorch backend that scales neural reconstructions to city-scale scenes with over 1 billion Gaussian splats, exceeding current benchmarks by more than 25 times.
Quick Take
This paper introduces a multi-GPU Gaussian splatting method using a PyTorch backend that scales neural reconstructions to city-scale scenes with over 1 billion Gaussian splats, exceeding current benchmarks by more than 25 times. The approach simplifies model distribution across GPUs without explicit cross-device communication, leveraging CUDA unified memory and NVLink.
Key Points
- Scalable Gaussian splatting for high-resolution neural reconstructions.
- Utilizes CUDA unified memory and NVLink for efficient GPU distribution.
- Achieves over 1 billion Gaussian splats in city-scale reconstructions.
- Model code requires no explicit cross-device communication.
- Exceeds current state-of-the-art by more than 25 times.
Paper Resources
Article Excerpt
From source RSS / original summaryarXiv:2606. 11390v1 Announce Type: new Abstract: Gaussian splatting methods have become increasingly popular for neural reconstruction of the real world. However, they are often limited in scale and resolution due to compute and memory constraints. We present a multi-GPU Gaussian splatting approach that scales reconstruction to higher resolutions and larger scenes while abstracting away the code complexity typically associated with distributing a model.
To accomplish this, we propose a PyTorch backend that distributes the Gaussian parameters and splatting operators across GPUs via CUDA unified memory and NVLink. Because distribution occurs at the operator level, the model code requires no explicit cross-device communication. More broadly, the backend exposes multiple GPUs as an aggregate PyTorch device and supports other PyTorch operators.
We demonstrate city-scale reconstructions with street-level detail consisting of over 1 billion Gaussian splats, more than 25 times as many as the current state of the art.
Reader Mode unavailable (could not extract clean content).
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CV
See more →LLM-Guided ANN Index Optimization for Human-Object Interaction Retrieval
A phase-aware LLM agent optimizes human-object interaction retrieval, outperforming Optuna TPE by 33.3% and VDTuner by 34.2% on the HICO-DET benchmark. This method enhances throughput by 15.3x over UniIR and demonstrates strong transferability across vector database management systems.
