On The Fairness Impacts of Hardware Selection in Machine Learning

TL;DR

This paper explores how hardware choices in machine learning can influence model fairness and performance disparities, revealing underlying factors and proposing mitigation strategies to address hardware-induced biases.

Contribution

It provides a combined theoretical and empirical analysis of hardware impacts on fairness, highlighting overlooked disparities and offering mitigation techniques.

Findings

Hardware selection affects model fairness and performance.

Differences in gradient flows contribute to disparities.

Proposed strategies can mitigate hardware-induced biases.

Abstract

In the machine learning ecosystem, hardware selection is often regarded as a mere utility, overshadowed by the spotlight on algorithms and data. This oversight is particularly problematic in contexts like ML-as-a-service platforms, where users often lack control over the hardware used for model deployment. How does the choice of hardware impact generalization properties? This paper investigates the influence of hardware on the delicate balance between model performance and fairness. We demonstrate that hardware choices can exacerbate existing disparities, attributing these discrepancies to variations in gradient flows and loss surfaces across different demographic groups. Through both theoretical and empirical analysis, the paper not only identifies the underlying factors but also proposes an effective strategy for mitigating hardware-induced performance imbalances.