2015
DOI: 10.1145/2687651
|View full text |Cite
|
Sign up to set email alerts
|

Making the Most of SMT in HPC

Abstract: This work presents an end-to-end methodology for quantifying the performance and power benefits of simultaneous multithreading (SMT) for HPC centers and applies this methodology to a production system and workload. Ultimately, SMT's value system-wide depends on whether users effectively employ SMT at the application level. However, predicting SMT's benefit for HPC applications is challenging; by doubling the number of threads, the application's characteristics may change. This work proposes statistical modelin… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1

Citation Types

0
1
0

Year Published

2016
2016
2021
2021

Publication Types

Select...
3
2
1

Relationship

0
6

Authors

Journals

citations
Cited by 13 publications
(1 citation statement)
references
References 49 publications
0
1
0
Order By: Relevance
“…While SMT-based multicore systems are emerging as the norm for achieving the computing power necessary [199,200], it is equally important to map these application to maximise energy efficiency [34,201]. With energy efficient computing emerging as an important paradigm, recent approaches adopted to using linear programming or heuristics to map threads-to-cores based on a quantifiable relative gain metric.…”
Section: Energy Efficient Schedulingmentioning
confidence: 99%
“…While SMT-based multicore systems are emerging as the norm for achieving the computing power necessary [199,200], it is equally important to map these application to maximise energy efficiency [34,201]. With energy efficient computing emerging as an important paradigm, recent approaches adopted to using linear programming or heuristics to map threads-to-cores based on a quantifiable relative gain metric.…”
Section: Energy Efficient Schedulingmentioning
confidence: 99%