Monitoring Collective Communication Among GPUs

TL;DR

This paper introduces an extension of the ComScribe tool to detect and visualize collective and P2P GPU communications in NVIDIA's NCCL library, aiding performance analysis in HPC and AI workloads.

Contribution

It extends prior GPU communication detection tools to include collective operations, providing detailed data transfer metrics and visualizations for better performance tuning.

Findings

Successfully identified collective and P2P communications in GPU applications.

Generated communication matrices revealing data transfer patterns.

Demonstrated the tool on machine translation and image classification applications.

Abstract

Communication among devices in multi-GPU systems plays an important role in terms of performance and scalability. In order to optimize an application, programmers need to know the type and amount of the communication happening among GPUs. Although there are prior works to gather this information in MPI applications on distributed systems and multi-threaded applications on shared memory systems, there is no tool that identifies communication among GPUs. Our prior work, ComScribe, presents a point-to-point (P2P) communication detection tool for GPUs sharing a common host. In this work, we extend ComScribe to identify communication among GPUs for collective and P2P communication primitives in NVIDIA's NCCL library. In addition to P2P communications, collective communications are commonly used in HPC and AI workloads thus it is important to monitor the induced data movement due to collectives. Our tool extracts the size and the frequency of data transfers in an application and visualizes them as a communication matrix. To demonstrate the tool in action, we present communication matrices and some statistics for two applications coming from machine translation and image classification domains.