Programming Bare-Metal Accelerators with Heterogeneous Threading Models: A Case Study of Matrix-3000

TL;DR

This paper presents the development of a programming model, compiler, and libraries for the Matrix-3000 system, enabling effective programming of its complex heterogeneous accelerators for exascale supercomputing.

Contribution

It introduces a new software stack with low-level and high-level programming interfaces tailored for Matrix-3000's complex architecture, including an OpenCL compiler.

Findings

Successfully deployed on an exascale prototype system.

Provided native support for bare-metal accelerators.

Facilitated programming with a new software stack.

Abstract

As the hardware industry moves towards using specialized heterogeneous many-cores to avoid the effects of the power wall, software developers are finding it hard to deal with the complexity of these systems. This article shares our experience when developing a programming model and its supporting compiler and libraries for Matrix-3000, which is designed for next-generation exascale supercomputers but has a complex memory hierarchy and processor organization. To assist its software development, we developed a software stack from scratch that includes a low-level programming interface and a high-level OpenCL compiler. Our low-level programming model offers native programming support for using the bare-metal accelerators of Matrix-3000, while the high-level model allows programmers to use the OpenCL programming standard. We detail our design choices and highlight the lessons learned from developing systems software to enable the programming of bare-metal accelerators. Our programming models have been deployed to the production environment of an exascale prototype system.