Performance Analysis of Scientific Applications on an NVIDIA Grace System

The NSF funded Frontera system has supported many important scientific applications over the last five years, not only in HPC but also in the area of Big Data and Machine Learning. Applications leveraged the petascale computational capabilities of the system to achieve new breakthroughs in science a...

Full description

Saved in:
Bibliographic Details
Published in:SC24-W: Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis pp. 558 - 566
Main Authors: Ruhela, Amit, Cazes, John, McCalpin, John, Del-Castillo-Negrete, Carlos, Li, Junjie, Liu, Hang, Chen, Hanning, Lu, Chun-Yaung, Milfeld, Kent, Zhang, Wenyang, Wang, Ian, Koesterke, Lars, DeSantis, John, Lewis, Nic, Hempel, Sean, Stanzione, Dan
Format: Conference Proceeding
Language:English
Published: IEEE 17.11.2024
Subjects:
Computer architecture
Grace-Hopper
Graphics processing units
Hardware
High performance computing
HPC
Lakes
Machine learning
NVIDIA
Performance analysis
Petascale
Production
Software
Supercomputers
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The NSF funded Frontera system has supported many important scientific applications over the last five years, not only in HPC but also in the area of Big Data and Machine Learning. Applications leveraged the petascale computational capabilities of the system to achieve new breakthroughs in science and engineering. Frontera is currently the fastest academic supercomputer in the US and is used by hundreds of scientists and engineers daily, enabling researchers to tackle more significant and complex challenges than ever before. As Frontera is nearing the end of its production life, it will be replaced by a new system, "Horizon". An intermediate system, "Vista", will bridge the gap by enabling researchers to access updated software and hardware technologies before Horizon is available. This paper summarizes early experiences on Vista by reporting on the performance of key applications and presenting its design and architecture. Early results are presented from the CPU architecture of Vista, "Grace-Grace", and compared with other processor technologies at TACC from Intel and AMD. A future paper will discuss the performance of the "Grace-Hopper" component of Vista.
DOI:10.1109/SCW63240.2024.00078