Runtime Support for Performance Portability on Heterogeneous Distributed Platforms

TL;DR

This paper presents a runtime framework that simplifies programming for heterogeneous distributed systems, achieving significant performance improvements and scalability across diverse hardware configurations.

Contribution

It introduces a scalable runtime system that provides portable abstractions for heterogeneous hardware, enabling efficient performance portability and inter-node communication.

Findings

Up to 300% performance improvement on a single device.

Linear scalability on systems with four GPUs.

Performance gains of up to 20% over MPI+CUDA for large messages.

Abstract

Hardware heterogeneity is here to stay for high-performance computing. Large-scale systems are currently equipped with multiple GPU accelerators per compute node and are expected to incorporate more specialized hardware. This shift in the computing ecosystem offers many opportunities for performance improvement; however, it also increases the complexity of programming for such architectures. This work introduces a runtime framework that enables effortless programming for heterogeneous systems while efficiently utilizing hardware resources. The framework is integrated within a distributed and scalable runtime system to facilitate performance portability across heterogeneous nodes. Along with the design, this paper describes the implementation and optimizations performed, achieving up to 300% improvement on a single device and linear scalability on a node equipped with four GPUs. The framework in a distributed memory environment offers portable abstractions that enable efficient inter-node communication among devices with varying capabilities. It delivers superior performance compared to MPI+CUDA by up to 20% for large messages while keeping the overheads for small messages within 10\%. Furthermore, the results of our performance evaluation in a distributed Jacobi proxy application demonstrate that our software imposes minimal overhead and achieves a performance improvement of up to 40%. This is accomplished by the optimizations at the library level as well as by creating opportunities to leverage application-specific optimizations like over-decomposition.