Communications of the ACM
Streamlining Accelerated Computing For Industry
Imperial College London researchers have developed software to enable computational fluid dynamics applications to make effective use of supercomputers.
Credit: Imperial College
Researchers from Imperial College London have developed open source software that will enable computational fluid dynamics (CFD) applications to make good use of supercomputers that have increased in size and scale.
They say PyFR, a Python-based application, combines highly accurate numerical methods with a highly flexible, portable, and scalable code implementation that makes efficient use of accelerators such as the graphical processing units (GPUs) of the Titan supercomputer at the Oak Ridge Leadership Computing Facility.
The researchers say scientists will be able to run PyFR on large-scale accelerated architectures to accurately resolve fluid features that were previously out of reach.
The team demonstrated the high-performance computing prowess of PyFR by running a high-resolution, GPU-accelerated simulation of flow over a jet turbine linear cascade on Titan. On the supercomputer, the team's highest-performing run contained 195 billion degrees of freedom--or independent variables--and operated at a sustained speed of 13.7 petaflops.
The team says the code could enable companies to better exploit petascale computing to understand long-standing fluid flow problems.
In recognition of the work, the team has been named a 2016 finalist for the ACM Gordon Bell Prize, the winner of which will be announced in November during SC16 in Salt Lake City, UT.
From Oak Ridge National Laboratory
View Full Article
Abstracts Copyright © 2016 Information Inc., Bethesda, Maryland, USA
No entries found