Leonardo Dagum

OpenMP: an industry standard API for shared-memory programming

At its most elemental level, OpenMP is a set of compiler directives and callable runtime library routines that extend Fortran (and separately, C and C++ to express shared memory parallelism. It leaves the base language unspecified, and vendors can implement OpenMP in any Fortran compiler. Naturally, to support pointers and allocatables, Fortran 90 and Fortran 95 require the OpenMP implementation to include additional semantics over Fortran 77. OpenMP leverages many of the X3H5 concepts while extending them to support coarse grain parallelism. The standard also includes a callable runtime library with accompanying environment variables.

The NAS parallel benchmarks summary and preliminary results

No abstract available

NAS parallel benchmark results

Benchmark results for the Numerical Aerodynamic Simulation (NAS) Program at NASA Ames Research Center, which is dedicated to advancing the science of computational aerodynamics are presented. The benchmark performance results are for the Y-MP, Y-MO EL, and C-90 systems from Cray Research; the TC2000 from Bolt Baranek and Newman; the Gamma iPSC/860 from Intel; the CM-2, CM-200, and CM-5 from Thinking Machines; the CS-1 from Meiko Scientific; the MP-1 and MP-2 from MasPar Computer; and the KSR-1 from Kendall Square Research. The results for the MP-1 and -2, the KSR-1, and the CM-5 have not been published before. Many of the other results are improved from previous listings, reflecting improvements both in compilers and in implementations.<<ETX>>

The NAS parallel benchmarks

No abstract available

Data parallel sorting for particle simulation

Sorting on a parallel architecture is a communications intensive event which can incur a high penalty in applications where it is required. In the case of particle simulation, only integer sorting is necessary, and sequential implementations easily attain the minimum performance bound of O (N) for N particles. Parallel implementations, however, have to cope with the parallel sorting problem which, in addition to incurring a heavy communications cost, can make the minimun performance bound difficult to attain. This paper demonstrates how the sorting problem in a particle simulation can be reduced to a merging problem, and describes an efficient data parallel algorithm to solve this merging problem in a particle simulation. The new algorithm is shown to be optimal under conditions usual for particle simulation, and its fieldwise implementation on the Connection Machine is analyzed in detail. The new algorithm is about four times faster than a fieldwise implementation of radix sort on the Connection Machine.

Experience in using SIMD and MIMD parallelism for computational fluid dynamics

One of the key objectives of the Applied Research Branch in the Numerical Aerodynamic Simulation (NAS) Systems Division at NASA Ames Research Center is the accelerated introduction of highly parallel machines into a fully operational environment. In this report we summarize some of the experiences with the parallel testbed machines at the NAS Applied Research Branch. We discuss the performance results obtained from the implementation of two computational fluid dynamics (CFD) applications, an unstructured grid solver and a particle simulation, on the Connection Machine CM-2 and the Intel iPSC/860.

Automatic partitioning of unstructured grids into connected components

This paper presents two partitioning schemes that guarantee connected components given a connected initial grid. Connected components are important for convergence of methods such as domain decomposition or multigrid. For many of the grids tested, the schemes produce partitions as good (in terms of number of cut edges) or better than spectral partitioning and require only modest computational resources. This paper describes the two schemes in detail and presents comparison results from a number of two and three dimensional unstructured grids.

Particle simulation on heterogeneous distributed supercomputers

The authors describe the implementation and performance of a three dimensional particle simulation distributed between a Thinking Machines CM-2 and a Cray Y-MP. These are connected by a combination of two high-speed networks; a high performance parallel interface (HIPPI) and an optical network (Ultra Net). This is the first application to use this configuration at NASA Ames Research Center. The authors describe their experience implementing and using the application and report the results of several timing measurements. They show that the distribution of applications across disparate supercomputing platforms is feasible and has reasonable performance. In addition, several practical aspects of the computing environment are discussed.<<ETX>>

The NAS parallel benchmarks—summary and preliminary results

No abstract available