A Simulation Platform for Multi-tenant Machine Learning Services on Thousands of GPUs
This work addresses the problem of optimizing resource management for data centers running large-scale ML services, though it is incremental as it builds on existing simulation and scheduling concepts.
The paper tackles the challenge of evaluating multi-tenant machine learning services on large GPU clusters by introducing AnalySIM, a simulation platform that enables efficient testing of scheduling policies, resulting in findings that preemption and migration reduce average job completion time and mitigate resource fragmentation.
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.