Study of Automatic GPU Offloading Method from Various Language Applications

TL;DR

This paper explores a universal method for automatic GPU offloading across multiple programming languages, aiming to simplify heterogeneous hardware utilization without requiring extensive technical skills.

Contribution

It introduces a environment-adaptive software approach that enables automatic offloading and high performance operation for applications written in C, Python, and Java.

Findings

Proposes a common offloading method for multiple languages.

Demonstrates automatic GPU offloading for C, Python, and Java applications.

Enhances accessibility of heterogeneous hardware computing.

Abstract

In recent years, utilization of heterogeneous hardware other than small core CPU such as GPU, FPGA or many core CPU is increasing. However, when using heterogeneous hardware, barriers of technical skills such as CUDA are high. Based on that, I have proposed environment-adaptive software that enables automatic conversion, configuration, and high performance operation of once written code, according to the hardware to be placed. However, the source language for offloading was mainly C/C++ language applications currently, and there was no research for common offloading for various language applications. In this paper, I study a common method for automatically offloading for various language applications not only in C language but also in Python and Java.