Cover Image
QR Code

Keeping Science on Keel When Software Moves.

The article examines the problem of false or misleading conclusions in science that stem from difficulties in reproducibility associated with porting more efficient scientific software in high performance computing (HPC). It discusses the concept of bisection, statistical combinations testing in the...

Full description

Published in:Communications of the ACM Vol. 64; no. 2; pp. 66 - 75
Main Authors: AHN, DONG H., BAKER, ALLISON H., BENTLEY, MICHAEL, BRIGGS, IAN, GOPALAKRISHNAN, GANESH, HAMMERLING, DORIT M., LAGUNA, IGNACIO, LEE, GREGORY L., MILROY, DANIEL J., VERTENSTEIN, MARIANA
Format: Article
Published: Association for Computing Machinery Feb2021
Subjects:
High performance computing
Computer software development
Application software porting
Reproducible research
Science
Engineering
Online Access:View this record in EBSCOhost
Summary:The article examines the problem of false or misleading conclusions in science that stem from difficulties in reproducibility associated with porting more efficient scientific software in high performance computing (HPC). It discusses the concept of bisection, statistical combinations testing in the Community Earth System Model (CESM), and the FLiT workflow.