An analysis is presented of the primary factors influencing the performance of a parallel implementation of the UCLA atmospheric general circulation model (AGCM) on distributedmemory, massively parallel computer systems. Several modifications to the original parallel AGCM code aimed at improving its
Reducing communication overhead: A parallel code optimization
✍ Scribed by Duane A. Bailey; Janice E. Cuny; Bruce B. MacLeod
- Publisher
- Elsevier Science
- Year
- 1987
- Tongue
- English
- Weight
- 872 KB
- Volume
- 4
- Category
- Article
- ISSN
- 0743-7315
No coin nor oath required. For personal study only.
✦ Synopsis
In highly parallel computation, where a large number of closely coupled processors cooperate on a single task, communication overhead is often the dominant factor in efficiency. We report here on parallel code optimixations designed to reduce this overhead. Our optimizations are targeted for synchronous machines. Because we have found such machines to be difficult to program, we start with a high-level asynchronous code which we automatically convert for synchronous execution before applying our optimixations. Using relatively simple transformations, we are able to demonstrate significant performance improvements for many programs. CB 1987 AU+ demic Press, Inc.
📜 SIMILAR VOLUMES
In this work, we simulate the interaction between intense laser radiation and a fully ionized plasma by solving a Vlasov-Maxwell system using the ''Particle-In-Cell'' (PIC) method. This method provides a very detailed description of the plasma dynamics, but at the expense of large computer resources