Solving GPU Assisted Exascale Computing Challenges with High-Order Finite Difference Methods

  • Anjum, Omer (Principal investigator)

Project Details

Description

High performance computing (HPC) is the primary venue to bring breakthroughs in frontier scientific applications, be it from physics, economics, biology, astronomy, medicine, genomics, eHealth, advanced communication networks and so on. Many of those applications demand the processing of Exabytes of data at the rate of Exaflops(1000 petaflops/sec). To date we are just at 93 petaflops/sec with fastest super computer on earth \cite{top500}, reason being the non-trivial challenges that exists both in hardware and software.
In this work the applicant will address some of those fundamental challenges exploring GPUs and also proposing novel HW and SW solutions. The applican will also develop a full MHD solver for GPUs which is also among the frist few attempts in the world, contributing to breakthrough science, particularly in HPC and in astrophysics from domain science. In the perspective of Finland, this is a unique opportunity to enhance the HPC competency in the country.
Short titleAnjum Omer
StatusFinished
Effective start/end date01/09/201730/04/2019

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.