Speaker
Alexei Strelchenko
(FNAL)
Description
The global all-to-all communications in the Krylov subspace iterative methods is
one of the major performance-limiting factors on large-scale parallel machines.
In this report we give a brief overview of recent algorithmic approaches
to mitigate communication cost in the iterative solvers. We present several variants
of communication-optimized fermion matrix inverters implemented in the QUDA library.
Finally, we will discuss a few possible scenarios of utilizing exascale-ready algorithms
from the Trilinos framework for LQCD applications.
Primary author
Alexei Strelchenko
(FNAL)