A Component Architecture for High-Performance Scientific Computing
International Journal of High Performance Computing Applications
Coupled fusion simulation using the common component architecture
ICCS'05 Proceedings of the 5th international conference on Computational Science - Volume Part I
Scientific Programming - Complexity in Scalable Computing
Proceedings of the 2009 Workshop on Component-Based High Performance Computing
Extending the concept of component interfaces: experience with the Integrated Plasma Simulator
Proceedings of the 2009 Workshop on Component-Based High Performance Computing
Event-based parareal: A data-flow based implementation of parareal
Journal of Computational Physics
Hi-index | 0.00 |
Successful simulation of the complex physics that affect magnetically confined fusion plasma remains an important target milestone towards the development of viable fusion energy. Major advances in the underlying physics formulations, mathematical modeling, and computational tools and techniques are needed to enable a complete fusion simulation on the emerging class of large scalecapability parallel computers that are coming on-line in the next few years. Several pilot projects are currently being undertaken to explore different (partial) code integration and coupling problems, and possible solutions that may guide the larger integration endeavor. In this paper, we present the design and implementation details of one such project, a component based approach to couple existing codes to model the interaction between high power radio frequency (RF) electromagnetic waves, and magnetohydrodynamics (MHD) aspects of the burning plasma. The framework and component design utilize a light coupling approach based on high level view of constituent codes that facilitates rapid incorporation of new components into the integrated simulation framework. The work illustrates the viability of the light coupling approach to better understand physics and stand-alone computer code dependencies and interactions, as a precursor to a more tightly coupled integrated simulation environment.