Abstract
Beginning with the March 1998 release of the Penn State University/NCAR Mesoscale Model (MM5), and continuing through eight subsequent releases up to the present, the official version has run on distributed -memory (DM) parallel computers. Source translation and runtime library support minimize the impact of parallelization on the original model source code, with the result that the majority of code is line-for-line identical with the original version. Parallel performance and scaling are equivalent to earlier, hand-parallelized versions; the modifications have no effect when the code is compiled and run without the DM option. Supported computers include the IBM SP, Cray T3E, Fujitsu VPP, Compaq Alpha clusters, and clusters of PCs (so-called Beowulf clusters). The approach also is compatible with shared-memory parallel directives, allowing distributed-memory/shared-memory hybrid parallelization on distributed-memory clusters of symmetric multiprocessors.
Highlights
The Pennsylvania State/National Center for Atmospheric Research Mesoscale Model is a limited-area model of atmospheric systems, in its fifth generation, MM5 [6]
Two criteria are used to assess the effectiveness of a same-source parallelization: impact on source code and model performance
The number of source lines affected without FLIC or any other mechanism would be enormous and would prevent a same-source implementation of DM-parallelism in a pre-existing code
Summary
The Pennsylvania State/National Center for Atmospheric Research Mesoscale Model is a limited-area model of atmospheric systems, in its fifth generation, MM5 [6]. It was designed for vector and sharedmemory parallel architectures. Two earlier distributedmemory (DM) parallel versions of the model code were developed at Argonne National Laboratory – the Massively Parallel Mesoscale Model (MPMM) and a subsequent Fortran implementation, MM90. These were efficient, scalable, and more modular and dynamically configurable [3,10] than the source model. Results are evaluated in terms of impact on model source code as well as model performance and scaling
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
