Leveraging GPU Tensor Cores for Double Precision Euclidean Distance Calculations

Benoit Gallet, Michael Gowanlock

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Scopus citations

Abstract

Tensor cores (TCs) are a type of Application-Specific Integrated Circuit (ASIC) and are a recent addition to Graphics Processing Unit (GPU) architectures. As such, TCs are purposefully designed to greatly improve the performance of Matrix Multiply-Accumulate (MMA) operations. While TCs are heavily studied for machine learning and closely related fields, where their high efficiency is undeniable, MMA operations are not unique to these fields. More generally, any computation that can be expressed as MMA operations can leverage TCs, and potentially benefit from their higher computational throughput compared to other general-purpose cores, such as CUDA cores on Nvidia GPUs. In this paper, we propose the first double precision (FP64) Euclidean distance calculation algorithm, which is expressed as MMA operations to leverage TCs on Nvidia GPUs, rather than the more commonly used CUDA cores. To show that the Euclidean distance can be accelerated in a real-world application, we evaluate our proposed TC algorithm on the distance similarity self-join problem, as the most computationally intensive part of the algorithm consists of computing distances in a multi-dimensional space. We find that the performance gain from using the tensor core algorithm over the CUDA core algorithm depends weakly on the dataset size and distribution, but is strongly dependent on data dimensionality. Overall, TCs are a compelling alternative to CUDA cores, particularly when the data dimensionality is low (≤ 4), as we achieve an average speedup of 1.28× and up to 2.23× against a state-of-the-art GPU distance similarity self-join algorithm. Furthermore, because this paper is among the first to explore the use of TCs for FP64 general-purpose computation, future research is promising.

Original languageEnglish (US)
Title of host publicationProceedings - 2022 IEEE 29th International Conference on High Performance Computing, Data, and Analytics, HiPC 2022
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages135-144
Number of pages10
ISBN (Electronic)9781665494236
DOIs
StatePublished - 2022
Externally publishedYes
Event29th Annual IEEE International Conference on High Performance Computing, Data, and Analytics, HiPC 2022 - Bangalore, India
Duration: Dec 18 2022Dec 21 2022

Publication series

NameProceedings - 2022 IEEE 29th International Conference on High Performance Computing, Data, and Analytics, HiPC 2022

Conference

Conference29th Annual IEEE International Conference on High Performance Computing, Data, and Analytics, HiPC 2022
Country/TerritoryIndia
CityBangalore
Period12/18/2212/21/22

Keywords

  • Euclidean Distance
  • GPU
  • Similarity Searches
  • Tensor Cores

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications
  • Hardware and Architecture
  • Information Systems
  • Information Systems and Management
  • Control and Optimization

Fingerprint

Dive into the research topics of 'Leveraging GPU Tensor Cores for Double Precision Euclidean Distance Calculations'. Together they form a unique fingerprint.

Cite this