Fine-grained MoE Load Balancing with Linear Programming

TL;DR

This paper introduces a novel parallelization strategy using linear programming to achieve fine-grained load balancing in Mixture-of-Experts models, significantly improving training throughput and GPU load distribution.

Contribution

It presents a new method for fine-grained load balancing in MoE systems that ensures optimal load distribution in every micro-batch using efficient token scheduling.

Findings

MicroMoE improves training throughput by up to 47.6%.

It almost always achieves optimal load balance among GPUs.

The method outperforms existing systems in load balancing efficiency.

Abstract

Mixture-of-Experts (MoE) has emerged as a promising approach to scale up deep learning models due to its significant reduction in computational resources. However, the dynamic nature of MoE leads to load imbalance among experts, severely impacting training efficiency. While previous research has attempted to address the load balancing challenge, existing solutions either compromise model accuracy or introduce additional system overhead. As a result, they fail to achieve fine-grained load balancing, which is crucial to optimizing training efficiency. We propose a novel parallelization strategy to achieve fine-grained load balancing in MoE systems. Our system is capable of achieving optimal load balancing in every micro-batch through efficient token scheduling across GPUs. Our experimental results demonstrate that MicroMoE improves the end-to-end training throughput by up to 47.6% compared with the state-of-the-art system, and almost consistently achieves optimal load balance among GPUs.