Heterogeneous Computing Systems for Complex Scientific Discovery Workflows

With Moore's law progressively running out of steam, heterogeneous computing architectures have been powering the top supercomputers in the world for many years and are now finding broader adoption across the industry. The trend towards sustainable computing also requires domain-specific hetero...

Full description

Saved in:
Bibliographic Details
Main Authors: Hagleitner, Christoph, Diamantopoulos, Dionysios, Ringlein, Burkhard, Evangelinos, Constantinos, Johns, Charles, Chang, Rong N., D'Amora, Bruce, Kahle, James A., Sexton, James, Johnston, Michael, Pyzer-Knapp, Edward, Ward, Chris
Format: Conference Proceeding
Language:English
Subjects:
accelerators
Computer architecture
disaggregation
heterogeneous computing
Heterogeneous networks
High performance computing
Industries
Supercomputers
Systems architecture
workflow
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:With Moore's law progressively running out of steam, heterogeneous computing architectures have been powering the top supercomputers in the world for many years and are now finding broader adoption across the industry. The trend towards sustainable computing also requires domain-specific heterogeneous hardware architectures, which promise further gains in energy efficiency. At the same time, today's high performance computing applications have evolved from monolithic simulations in a single domain to multidisciplinary complex workflows. In this paper, we explore how these trends affect system design decisions and what this means for future computing system architectures.
ISSN:1558-1101
DOI:10.23919/DATE51398.2021.9474061