发表机构
Rice University; Johns Hopkins University(莱斯大学; 约翰斯·霍普金斯大学)
机构由 AI 辅助整理,请以论文原文为准。AI 中文总结
本文提出GPU加速的Bregman Douglas-Rachford分裂算法,针对三种输入格式的离散最优传输问题,通过硬件感知设计提升数值稳定性与运行速度,并在统一度量下超越八个基线求解器,达到最先进性能。
AI 中文摘要
我们提出了GPU加速的Bregman Douglas-Rachford分裂算法(BDRS),用于解决三种输入格式下的离散最优传输问题:显式代价矩阵、点云及其之间的地面代价、以及规则网格上的可分离代价。针对每种输入格式,我们提出了BDRS迭代的数学等价表示的硬件感知设计,以增强数值稳定性和经验运行时间。我们在同一设备上,将所提出的三种实现与文献中的八个GPU基线求解器进行了基准测试。我们证明了我们的BDRS实现在各自输入格式上达到了最先进的性能。据我们所知,这是首次使用统一最优性度量对GPU DOT求解器进行跨求解器研究。
英文摘要
We present GPU-accelerated Bregman Douglas--Rachford splitting algorithm (BDRS) for discrete optimal transport problem in three input formats: an explicit cost matrix, a point cloud with a ground cost between them, and a separable cost on a regular grid. For each input format, we propose hardware-aware designs of mathematically equivalent representations for the BDRS iterations to enhance numerical stability and empirical runtime. We benchmark the three proposed implementations against eight GPU baseline solvers from the literature on the same device. We demonstrate that our implementations of BDRS achieve state-of-the-art performance on their respective input formats. To the best of our knowledge, this is the first cross-solver study of GPU DOT solvers with a unified measure of optimality.
Comments40 pages, 7 figures, 3 tables