A parallelizing compiler for distributed memory parallel computers
A parallelizing compiler for distributed memory parallel computers
Object oriented parallel programming: experiments and results
Proceedings of the 1991 ACM/IEEE conference on Supercomputing
The DINO parallel programming language
Journal of Parallel and Distributed Computing
Parallel methods for solving nonlinear block bordered systems of equations
SIAM Journal on Scientific and Statistical Computing
Efficient language constructs for complex data parallelism on distributed memory multiprocessors
Efficient language constructs for complex data parallelism on distributed memory multiprocessors
Data-Parallel Programming on Multicomputers
IEEE Software
Data-Parallel Programming on MIMD Computers
IEEE Transactions on Parallel and Distributed Systems
PASM: A Partitionable SIMD/MIMD System for Image Processing and Pattern Recognition
IEEE Transactions on Computers
Visualization and sonification of methods
PAS '95 Proceedings of the First Aizu International Symposium on Parallel Algorithms/Architecture Synthesis
| Hi-index | 0.00 |
The goal of the research described in this article is to developflexible language constructs for writing large data parallelnumerical programs for distributed memory (multiple instructionmultiple data [MIMD]) multiprocessors. Previously, several modelshave been developed to support synchronization and communication.Models for global synchronization include single instructionmultiple data (SIMD), single program multiple data (SPMD), andsequential programs annotated with data distribution statements.The two primary models for communication include implicitcommunication based on shared memory and explicit communicationbased on messages. None of these models by themselves seemsufficient to permit the natural and efficient expression of thevariety of algorithms that occur in large scientific computations.In this article, we give an overview of a new language thatcombines many of these programming models in a clean manner. Thisis done in a modular fashion such that different models can becombined to support large programs. Within a module, the selectionof a model depends on the algorithm and its efficiencyrequirements. In this article, we give an overview of the languageand discuss some of the critical implementation details.