An Overview of Performance Portability in the Uintah Runtime System through the Use of Kokkos
- Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
- Univ. of Utah, Salt Lake City, UT (United States)
The current diversity in nodal parallel computer architectures is seen in machines based upon multicore CPUs, GPUs and the Intel Xeon Phi’s. Here, a class of approaches for enabling scalability of complex applications on such architectures is based upon Asynchronous Many Task software architectures such as that in the Uintah framework used for the parallel solution of solid and fluid mechanics problems. Uintah has both an applications layer with its own programming model and a separate runtime system. While Uintah scales well today, it is necessary to address nodal performance portability in order for it to continue to do. Incrementally modifying Uintah to use the Kokkos performance portability library through prototyping experiments results in improved kernel performance by more than a factor of two.
- Research Organization:
- Univ. of Utah, Salt Lake City, UT (United States)
- Sponsoring Organization:
- USDOE National Nuclear Security Administration (NNSA)
- DOE Contract Number:
- NA0002375; AC04-94AL85000
- OSTI ID:
- 1756094
- Journal Information:
- 2016 Second International Workshop on Extreme Scale Programming Models and Middlewar (ESPM2), Vol. 1; Conference: 2016 Second International Workshop on Extreme Scale Programming Models and Middlewar (ESPM2), Salt Lake City, UT (United States), 18 Nov 2016
- Country of Publication:
- United States
- Language:
- English
Evaluation of performance portability frameworks for the implementation of a particle‐in‐cell code
|
journal | December 2019 |
A High-performance and Portable All-Mach Regime Flow Solver Code with Well-balanced Gravity. Application to Compressible Convection
|
journal | April 2019 |
Similar Records
Demonstrating GPU code portability and scalability for radiative heat transfer computations
Improving Uintah's Scalability Through the Use of Portable Kokkos-Based Data Parallel Tasks