MT4G: A Tool for Reliable Auto-Discovery of NVIDIA and AMD GPU Compute and Memory Topologies
Stepan Vanecek, Manuel Walter Mussbacher, Dominik Groessler, Urvij Saroliya, and Martin Schulz

TL;DR
MT4G is an open-source, vendor-agnostic tool that automatically discovers detailed GPU topologies and configurations, aiding performance optimization in HPC and AI applications.
Contribution
It introduces a novel, automated method combining APIs and microbenchmarks with statistical analysis to reliably identify GPU topological attributes across vendors.
Findings
Successfully applied to ten different GPUs
Enhanced GPU performance modeling accuracy
Improved resource management in workflows
Abstract
Understanding GPU topology is essential for performance-related tasks in HPC or AI. Yet, unlike for CPUs with tools like hwloc, GPU information is hard to come by, incomplete, and vendor-specific. In this work, we address this gap and present MT4G, an open-source and vendor-agnostic tool that automatically discovers GPU compute and memory topologies and configurations, including cache sizes, bandwidths, and physical layouts. MT4G combines existing APIs with a suite of over 50 microbenchmarks, applying statistical methods, such as the Kolmogorov-Smirnov test, to automatically and reliably identify otherwise programmatically unavailable topological attributes. We showcase MT4G's universality on ten different GPUs and demonstrate its impact through integration into three workflows: GPU performance modeling, GPUscout bottleneck analysis, and dynamic resource partitioning. These…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsParallel Computing and Optimization Techniques · Big Data and Digital Economy · Cloud Computing and Resource Management
