ICL Research Profile



The Asynchronous Iterative Solvers for Extreme-scale Computing (AsyncIS) project aims to replace the outer Krylov subspace solver with an asynchronous optimized Schwarz method, thereby removing the global synchronization and bulk synchronous requirement in the execution of numerical codes. AsyncIS also proposes to develop and analyze asynchronous preconditioners (i.e., preconditioners that are generated and/or applied in asynchronous fashion).

The AsyncIS project is a collaboration between Georgia Tech, ICL/UTK, Temple University, and Sandia National Laboratories, and is funded by DOE to address the massive amounts of parallelism that will characterize extreme-scale supercomputers, such as an anticipated exascale machine.

In Collaboration With

  1. Georgia Tech University
  2. Sandia National Laboratories

Sponsored by

  1. The United States Department of Energy