Compiling Communication-eecient Programs for Massively Parallel Machines. Ieee on Compiling Array Expressions for Eecient Execution on Distributed Memory Machines. in 5.4 Compiler Optimizations 5.3 Multigrid Code 5.2 Multiblock Code

26] Max Lemke and Daniel Quinlan. P++, a C++ virtual shared grids based programming environment for architecture-independent development of structured grid applications.grid representation of emission source clusters in regional air quality modeling. 31] R. Meakin. Moving body overset grid methods for complete aircraft tiltrotor simulations, AIAA-93-3350. A multiscale nite element pollutant transport scheme for urban and regional modeling. The improved robustness of multigrid elliptic solvers based on multiple semicoarsened grids. 37] James M. Stichnoth. EEcient compilation of array statements for private memory multicomputers.2d: A radiation magnetohydrodynamics code for astrophysical ows in two space dimensions: I. the hydrodynamic algorithms and tests. Development of a exible and eecient multigrid-based multiblock ow solver; AIAA-93-0677. Distributed memory compiler methods for irregular problems-data copy reuse and runtime partitioning. 14] Survey of principal investigators of grand challenge applications: Workshop on grand challenge applications and software technology, May 1993. Global land cover classiication by remote sensing:present capabilities and future possibilities. 28 such an application by hand, it is very diicult to analyze the exact communication required and then to be able to use the communication routines provided by the machine to communicate eeciently. Our runtime and compiler support can be used to parallelize such applications conveniently. Our experimental results have shown that the code parallelized by using the compiler will have only a small overhead as compared to the best hand parallelized code (i.e. parallelized by invoking system's communication primitives by hand). While the design of our runtime system was motivated by multiblock and multigrid applications, our runtime primitives can be used in many cases for regular codes as well. We therefore, believe that our runtime support and compiler techniques can be used by compilers of HPF-style parallel programming languages in general. and Zeki Bozkus for many enlightening discussions and for allowing us to integrate our runtime support into their emerging Fortran 90D compiler. The detailed discussions we had with Sanjay Ranka, Alok Choudhary and Zeki Bozkus during their visits to Maryland were extremely productive. We are also grateful to V. Vatsa and M. Senetrik at NASA Langley for giving us access to the multiblock TLNS3D application code. We will also like to thank John van Rosendale at ICASE and Andrea Overman at NASA Langley for making their sequential and hand parallelized multigrid code available to us. We also thank Jim Humphries for creating a portable version of the runtime library.erating local addresses and communication sets for …