Saved in:
Bibliographic Details
Main Authors: Singhania, Varsha, Aga, Shaizeen, Ibrahim, Mohamed Assem
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2412.12426
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909556685144064
author Singhania, Varsha
Aga, Shaizeen
Ibrahim, Mohamed Assem
author_facet Singhania, Varsha
Aga, Shaizeen
Ibrahim, Mohamed Assem
contents Ubiquity of AI makes optimizing GPU power a priority as large GPU-based clusters are often employed to train and serve AI models. An important first step in optimizing GPU power consumption is high-fidelity and fine-grain power measurement of key AI computations on GPUs. To this end, we observe that as GPUs get more powerful, the resulting sub-millisecond to millisecond executions make fine-grain power analysis challenging. In this work, we first carefully identify the challenges in obtaining fine-grain GPU power profiles. To address these challenges, we devise FinGraV methodology where we employ execution time binning, careful CPU-GPU time synchronization, and power profile differentiation to collect fine-grain GPU power profiles across prominent AI computations and across a spectrum of scenarios. Using the said FinGraV power profiles, we provide both, guidance on accurate power measurement and, in-depth view of power consumption on state-of-the-art AMD Instinct MI300X. For the former, we highlight a methodology for power differentiation across executions. For the latter, we make several observations pertaining to GPU sub-component power consumption and GPU power proportionality across different scenarios. We believe that FinGraV unlocks both an accurate and a deeper view of power consumption of GPUs and opens up avenues for power optimization of these ubiquitous accelerators.
format Preprint
id arxiv_https___arxiv_org_abs_2412_12426
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle FinGraV: Methodology for Fine-Grain GPU Power Visibility and Insights
Singhania, Varsha
Aga, Shaizeen
Ibrahim, Mohamed Assem
Hardware Architecture
Distributed, Parallel, and Cluster Computing
Ubiquity of AI makes optimizing GPU power a priority as large GPU-based clusters are often employed to train and serve AI models. An important first step in optimizing GPU power consumption is high-fidelity and fine-grain power measurement of key AI computations on GPUs. To this end, we observe that as GPUs get more powerful, the resulting sub-millisecond to millisecond executions make fine-grain power analysis challenging. In this work, we first carefully identify the challenges in obtaining fine-grain GPU power profiles. To address these challenges, we devise FinGraV methodology where we employ execution time binning, careful CPU-GPU time synchronization, and power profile differentiation to collect fine-grain GPU power profiles across prominent AI computations and across a spectrum of scenarios. Using the said FinGraV power profiles, we provide both, guidance on accurate power measurement and, in-depth view of power consumption on state-of-the-art AMD Instinct MI300X. For the former, we highlight a methodology for power differentiation across executions. For the latter, we make several observations pertaining to GPU sub-component power consumption and GPU power proportionality across different scenarios. We believe that FinGraV unlocks both an accurate and a deeper view of power consumption of GPUs and opens up avenues for power optimization of these ubiquitous accelerators.
title FinGraV: Methodology for Fine-Grain GPU Power Visibility and Insights
topic Hardware Architecture
Distributed, Parallel, and Cluster Computing
url https://arxiv.org/abs/2412.12426