Languages for vector and parallel processors
β Scribed by R.H. Perrott
- Publisher
- Elsevier Science
- Year
- 1982
- Tongue
- English
- Weight
- 800 KB
- Volume
- 26
- Category
- Article
- ISSN
- 0010-4655
No coin nor oath required. For personal study only.
β¦ Synopsis
This paper first considers the major developments which have occurred in the design of high level languages for sequential machines. These developments illustrate how languages which were independent of the hardware eventually evolved. Two major types of language for vector and parallel processors have evolved, namely, detection of parallelism languages and expression of machine parallelism languages. The disadvantages and advantages of each type of language are examined. A third type of language is also considered which reflects neither the compiler's detection mechanism nor the underlying hardware. The syntax of this language enables the parallel nature of a problem to be expressed directly. The language is thus appropriate for both vector and array processors.
π SIMILAR VOLUMES
We introduce an algorithm, called TMO (Two-Machine Optimal Scheduling) which minimizes the makespan for two identical processors. TMO employs lexicographic search in conjunction with the longest-processing time sequence to derive an optimal schedule. For the m identical parallel processors problem,
## Abstract We describe two algorithms for the parallel calculation of a CHARMmβlike force field in macromolecules. For a molecule with a given number of atoms, we show that there is an optimal number of processors leading to a minimum computation time. At the optimum, both the number of processors
We investigate and test an algorithm suitable for the parallel calculation of the potential energy of a protein, or its spatial gradient, when the protein atoms interact via pair potentials. This algorithm is similar to one previously proposed, but it is more efficient, having half the interprocesso