Understanding Layered Portability from HPC to Cloud in Containerized Environments

TL;DR

This paper investigates how different layers in containerized environments affect the performance of HPC applications migrating from dedicated HPC systems to cloud platforms, focusing on ARM-based hardware including Nvidia Grace CPU.

Contribution

It provides a detailed characterization of the performance impact of container virtualization layers, host OS, kernel, and execution modes on HPC applications across diverse ARM-based platforms.

Findings

Container overhead is less than 4% for some applications.

Overhead increases to 8-10% for others like FFT and HPCG.

Changing container execution modes has negligible performance impact.

Abstract

Recent development in lightweight OS-level virtualization, containers, provides a potential solution for running HPC applications on the cloud platform. In this work, we focus on the impact of different layers in a containerized environment when migrating HPC containers from a dedicated HPC system to a cloud platform. On three ARM-based platforms, including the latest Nvidia Grace CPU, we use six representative HPC applications to characterize the impact of container virtualization, host OS and kernel, and rootless and privileged container execution. Our results indicate less than 4\% container overhead in DGEMM, miniMD, and XSBench, but 8\%-10\% overhead in FFT, HPCG, and Hypre. We also show that changing between the container execution modes results in negligible performance differences in the six applications.