HeAT -- a Distributed and GPU-accelerated Tensor Framework for Data Analytics

TL;DR

HeAT is a GPU-accelerated, distributed tensor framework with a NumPy-like API that enables large-scale data analysis on high-performance systems, significantly improving speed over similar tools.

Contribution

HeAT introduces a scalable, MPI-based tensor framework leveraging PyTorch for distributed data analysis with an easy-to-use interface.

Findings

Achieves up to 100x speedup compared to similar frameworks.

Supports large-scale parallel processing on HPC systems.

Provides both low-level and high-level array computations.

Abstract

To cope with the rapid growth in available data, the efficiency of data analysis and machine learning libraries has recently received increased attention. Although great advancements have been made in traditional array-based computations, most are limited by the resources available on a single computation node. Consequently, novel approaches must be made to exploit distributed resources, e.g. distributed memory architectures. To this end, we introduce HeAT, an array-based numerical programming framework for large-scale parallel processing with an easy-to-use NumPy-like API. HeAT utilizes PyTorch as a node-local eager execution engine and distributes the workload on arbitrarily large high-performance computing systems via MPI. It provides both low-level array computations, as well as assorted higher-level algorithms. With HeAT, it is possible for a NumPy user to take full advantage of their available resources, significantly lowering the barrier to distributed data analysis. When compared to similar frameworks, HeAT achieves speedups of up to two orders of magnitude.