Bridging Evolutionary Multiobjective Optimization and GPU Acceleration via Tensorization

2025-03-26

AI Summary

Traditional evolutionary multi-objective optimization (EMO) algorithms suffer from limited parallelism and poor hardware compatibility when tackling large-scale, complex problems. To address this, this paper introduces the first EMO tensorization paradigm, enabling efficient GPU-accelerated implementations of mainstream algorithms—including NSGA-III, MOEA/D, and HypE—via unified tensor-based semantic modeling. The approach integrates low-level CUDA acceleration, PyTorch tensor operations, and the NVIDIA PhysX physics engine to construct the first GPU-accelerated multi-objective robotic control benchmark. Experimental results demonstrate up to 1,113× speedup over CPU-based implementations, scalable support for population sizes exceeding one million, and preservation of Pareto solution quality—while simultaneously enhancing behavioral policy diversity and overall performance.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Intelligent Robots: Learning & Optimization for ROB
• Machine Learning: Matrix & Tensor Methods

Application Category

• Economics, Online Markets and Human Computation: Fairness, privacy, and diversity in economic environments
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Evolutionary multiobjective optimization (EMO) has made significant strides over the past two decades. However, as problem scales and complexities increase, traditional EMO algorithms face substantial performance limitations due to insufficient parallelism and scalability. While most work has focused on algorithm design to address these challenges, little attention has been given to hardware acceleration, thereby leaving a clear gap between EMO algorithms and advanced computing devices, such as GPUs. To bridge the gap, we propose to parallelize EMO algorithms on GPUs via the tensorization methodology. By employing tensorization, the data structures and operations of EMO algorithms are transformed into concise tensor representations, which seamlessly enables automatic utilization of GPU computing. We demonstrate the effectiveness of our approach by applying it to three representative EMO algorithms: NSGA-III, MOEA/D, and HypE. To comprehensively assess our methodology, we introduce a multiobjective robot control benchmark using a GPU-accelerated physics engine. Our experiments show that the tensorized EMO algorithms achieve speedups of up to 1113x compared to their CPU-based counterparts, while maintaining solution quality and effectively scaling population sizes to hundreds of thousands. Furthermore, the tensorized EMO algorithms efficiently tackle complex multiobjective robot control tasks, producing high-quality solutions with diverse behaviors. Source codes are available at https://github.com/EMI-Group/evomo.

Problem

• Bridging EMO algorithms and GPU acceleration via tensorization
• Addressing performance limitations in large-scale EMO problems
• Enabling scalable parallelism for complex multiobjective optimization

Innovation

• Parallelize EMO on GPUs via tensorization
• Transform EMO operations into tensor representations
• Achieve significant speedups with GPU acceleration