ATLAHS: An Application-centric Network Simulator Toolchain for AI, HPC, and Distributed Storage

TL;DR

ATLAHS is a versatile, open-source network simulation toolchain that accurately models real-world AI, HPC, and storage workloads, outperforming existing simulators in speed and trace efficiency.

Contribution

It introduces a flexible, application-centric simulation framework that supports multiple backends and scenarios, enabling more realistic and comprehensive performance analysis.

Findings

Achieves less than 5% error in workload simulation accuracy.

Outperforms AstraSim in runtime and trace size efficiency.

Demonstrates utility through case studies on congestion control and job placement.

Abstract

Network simulators play a crucial role in evaluating the performance of large-scale systems. However, existing simulators rely heavily on synthetic microbenchmarks or narrowly focus on specific domains, limiting their ability to provide comprehensive performance insights. In this work, we introduce ATLAHS, a flexible, extensible, and open-source toolchain designed to trace real-world applications and accurately simulate their workloads. ATLAHS leverages the GOAL format to model communication and computation patterns in AI, HPC, and distributed storage applications. It supports multiple network simulation backends and handles multi-job and multi-tenant scenarios. Through extensive validation, we demonstrate that ATLAHS achieves high accuracy in simulating realistic workloads (consistently less than 5% error), while significantly outperforming AstraSim, the current state-of-the-art AI systems simulator, in terms of simulation runtime and trace size efficiency. We further illustrate ATLAHS's utility via detailed case studies, highlighting the impact of congestion control algorithms on the performance of distributed storage systems, as well as the influence of job-placement strategies on application runtimes.