Similarity Search with Tensor Core Units

TL;DR

This paper demonstrates that Tensor Core Units can significantly accelerate similarity search tasks by adapting algorithms for dimensionality reduction and similarity join, achieving a speedup proportional to the hardware parameter m.

Contribution

The paper introduces novel algorithms leveraging TCUs for similarity search, providing a substantial speedup over traditional methods.

Findings

Achieves a  speedup using TCUs.

Demonstrates effective adaptation of Johnson-Lindenstrauss and similarity join algorithms for hardware acceleration.

Provides theoretical and empirical validation of the proposed methods.

Abstract

Tensor Core Units (TCUs) are hardware accelerators developed for deep neural networks, which efficiently support the multiplication of two dense $\sqrt{m}\times \sqrt{m}$ matrices, where $m$ is a given hardware parameter. In this paper, we show that TCUs can speed up similarity search problems as well. We propose algorithms for the Johnson-Lindenstrauss dimensionality reduction and for similarity join that, by leveraging TCUs, achieve a $\sqrt{m}$ speedup up with respect to traditional approaches.