Exploring Application Performance on Emerging Hybrid-Memory Supercomputers

TL;DR

This paper systematically evaluates how emerging hybrid-memory supercomputers impact application performance, showing benefits for data analytics and potential improvements for traditional scientific workloads at large scale.

Contribution

It introduces a methodology to analyze application performance on hybrid-memory systems, modeling memory as fast and slow tiers and comparing traditional and emerging workloads.

Findings

Data analytics applications benefit from hybrid-memory systems at large scale.

Traditional scientific applications do not suffer performance penalties and may improve.

Hybrid-memory systems can enhance overall supercomputer performance.

Abstract

Next-generation supercomputers will feature more hierarchical and heterogeneous memory systems with different memory technologies working side-by-side. A critical question is whether at large scale existing HPC applications and emerging data-analytics workloads will have performance improvement or degradation on these systems. We propose a systematic and fair methodology to identify the trend of application performance on emerging hybrid-memory systems. We model the memory system of next-generation supercomputers as a combination of "fast" and "slow" memories. We then analyze performance and dynamic execution characteristics of a variety of workloads, from traditional scientific applications to emerging data analytics to compare traditional and hybrid-memory systems. Our results show that data analytics applications can clearly benefit from the new system design, especially at large scale. Moreover, hybrid-memory systems do not penalize traditional scientific applications, which may also show performance improvement.

Full text

See pages 1-last of sbacpad.pdf