10 Years Later: Cloud Computing is Closing the Performance Gap

TL;DR

This paper investigates whether modern cloud computing can now match high-performance computing (HPC) performance for scientific applications, finding that recent cloud advancements enable HPC-competitive performance at modest scales.

Contribution

It introduces a multi-level approach to analyze the performance gap between HPC and cloud computing, highlighting recent cloud improvements that close this gap.

Findings

High-end cloud can deliver HPC-competitive performance for various applications

Recent cloud hardware improvements include high-speed memory and interconnects

Dedicated batch scheduling enhances cloud HPC performance at modest scales

Abstract

Can cloud computing infrastructures provide HPC-competitive performance for scientific applications broadly? Despite prolific related literature, this question remains open. Answers are crucial for designing future systems and democratizing high-performance computing. We present a multi-level approach to investigate the performance gap between HPC and cloud computing, isolating different variables that contribute to this gap. Our experiments are divided into (i) hardware and system microbenchmarks and (ii) user application proxies. The results show that today's high-end cloud computing can deliver HPC-competitive performance not only for computationally intensive applications but also for memory- and communication-intensive applications - at least at modest scales - thanks to the high-speed memory systems and interconnects and dedicated batch scheduling now available on some cloud platforms.