A Simulation Platform for Multi-tenant Machine Learning Services on Thousands of GPUs

TL;DR

AnalySIM is a trace-driven cluster simulator designed for efficient evaluation of multi-tenant machine learning services on thousands of GPUs, enabling policy testing without real cluster deployment.

Contribution

The paper introduces AnalySIM, a scalable simulation platform that models large GPU clusters for performance analysis of scheduling policies in multi-tenant ML workloads.

Findings

Preemption and migration reduce job completion time.

AnalySIM effectively models large GPU clusters.

Scheduling policies impact resource utilization.

Abstract

Multi-tenant machine learning services have become emerging data-intensive workloads in data centers with heavy usage of GPU resources. Due to the large scale, many tuning parameters and heavy resource usage, it is usually impractical to evaluate and benchmark those machine learning services on real clusters. In this demonstration, we present AnalySIM, a cluster simulator that allows efficient design explorations for multi-tenant machine learning services. Specifically, by trace-driven cluster workload simulation, AnalySIM can easily test and analyze various scheduling policies in a number of performance metrics such as GPU resource utilization. AnalySIM simulates the cluster computational resource based on both physical topology and logical partition. The tool has been used in SenseTime to understand the impact of different scheduling policies with the trace from a real production cluster of over 1000 GPUs. We find that preemption and migration are able to significantly reduce average job completion time and mitigate the resource fragmentation problem.