TY - CHAP
T1 - Application of high-performance computing for peridynamics
AU - Diyaroglu, Cagan
AU - Vazic, Bozo
AU - Oterkus, Erkan
AU - Oterkus, Selda
PY - 2021/4/30
Y1 - 2021/4/30
N2 - Peridynamic (PD) codes, which are written by using any programming language, e.g., C, C++, FORTRAN, or Java, can be run in any available computing facility. This can be a desktop/laptop computer or a high performance computing (HPC) facility. Nowadays, desktops/laptops include many central processing units (CPUs) and relatedly many cores in their architectures. Moreover, graphical processing units (GPUs) of computers have been improved above and beyond the expectations. Either one of them or both of them can be benefited in order to solve large and complicated problems in personal desktops/laptops. This chapter presents how it is possible to improve PD codes and make them efficient for large-scale problems with the help of parallel programming procedures. Moreover, how simple modifications with very basic knowledge of parallel programming skills can lead to significant increase in our calculation speeds is demonstrated. Specifically, CPU- and GPU-based approaches are explained for parallelizing PD codes.
AB - Peridynamic (PD) codes, which are written by using any programming language, e.g., C, C++, FORTRAN, or Java, can be run in any available computing facility. This can be a desktop/laptop computer or a high performance computing (HPC) facility. Nowadays, desktops/laptops include many central processing units (CPUs) and relatedly many cores in their architectures. Moreover, graphical processing units (GPUs) of computers have been improved above and beyond the expectations. Either one of them or both of them can be benefited in order to solve large and complicated problems in personal desktops/laptops. This chapter presents how it is possible to improve PD codes and make them efficient for large-scale problems with the help of parallel programming procedures. Moreover, how simple modifications with very basic knowledge of parallel programming skills can lead to significant increase in our calculation speeds is demonstrated. Specifically, CPU- and GPU-based approaches are explained for parallelizing PD codes.
KW - central processing units
KW - graphical processing units
KW - high performance computing
KW - parallel programming
KW - peridynamics
U2 - 10.1016/B978-0-12-820069-8.00017-2
DO - 10.1016/B978-0-12-820069-8.00017-2
M3 - Chapter
SN - 9780128200698
T3 - A volume in Elsevier Series in Mechanics of Advanced Materials
SP - 403
EP - 417
BT - Peridynamic Modeling, Numerical Techniques, and Applications
A2 - Oterkus, Erkan
A2 - Oterkus, Selda
A2 - Madenci, Erdogan
CY - Amsterdam, Netherlands
ER -