CUTHERMO: Understanding GPU Memory Inefficiencies with Heat Map Profiling

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Zhao, Yanbo, Cui, Jinku, Li, Zecheng, Jiao, Shuyin, Liu, Xu, Li, Jiajia
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911075812769792
author Zhao, Yanbo
Cui, Jinku
Li, Zecheng
Jiao, Shuyin
Liu, Xu
Li, Jiajia
author_facet Zhao, Yanbo
Cui, Jinku
Li, Zecheng
Jiao, Shuyin
Liu, Xu
Li, Jiajia
contents GPUs have become indispensable in high-performance computing, machine learning, and many other domains. Efficiently utilizing the memory subsystem on GPUs is critical for maximizing computing power through massive parallelism. Analyzing memory access patterns has proven to be an effective method for understanding memory bottlenecks in applications. However, comprehensive runtime and fine-grained memory profiling support is lacking on GPU architectures. In this work, we introduce cuThermo, a lightweight and practical profiling tool for GPU memory analysis. It operates on GPU binaries without requiring any modifications to hardware, operating system, or application source code. Given a CUDA application, cuThermo identifies memory inefficiencies at runtime via a heat map based on distinct visited warp counts to represent word-sector-level data sharing and provides optimization guidance in performance tuning iterations. Through our experiments on six applications, we identified five memory access patterns that are portable across different GPU architectures. By evaluating optimization on two GPUs, cuThermo achieves up to $721.79\%$ performance improvement.
format Preprint
id arxiv_https___arxiv_org_abs_2507_18729
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CUTHERMO: Understanding GPU Memory Inefficiencies with Heat Map Profiling
Zhao, Yanbo
Cui, Jinku
Li, Zecheng
Jiao, Shuyin
Liu, Xu
Li, Jiajia
Distributed, Parallel, and Cluster Computing
Performance
GPUs have become indispensable in high-performance computing, machine learning, and many other domains. Efficiently utilizing the memory subsystem on GPUs is critical for maximizing computing power through massive parallelism. Analyzing memory access patterns has proven to be an effective method for understanding memory bottlenecks in applications. However, comprehensive runtime and fine-grained memory profiling support is lacking on GPU architectures. In this work, we introduce cuThermo, a lightweight and practical profiling tool for GPU memory analysis. It operates on GPU binaries without requiring any modifications to hardware, operating system, or application source code. Given a CUDA application, cuThermo identifies memory inefficiencies at runtime via a heat map based on distinct visited warp counts to represent word-sector-level data sharing and provides optimization guidance in performance tuning iterations. Through our experiments on six applications, we identified five memory access patterns that are portable across different GPU architectures. By evaluating optimization on two GPUs, cuThermo achieves up to $721.79\%$ performance improvement.
title CUTHERMO: Understanding GPU Memory Inefficiencies with Heat Map Profiling
topic Distributed, Parallel, and Cluster Computing
Performance
url https://arxiv.org/abs/2507.18729