Derivation of efficient parallel programs: An example from genetic sequence analysis
- Texas Univ., Austin, TX (USA) Dept. of Computer Sciences
- Argonne National Lab., IL (USA)
Implementation issues such as synchronization of shared data, implementation of abstract data types, and scheduling of processes are usually not addressed in the formal derivation of parallel programs. We seek to redress the situation by considering these issues in the context of developing an efficient implementation of an actual parallel program. The computational problem that we selected is motivated by work in aligning sequences of genetic material. We proceed by developing an algorithm in Unity and investigating the issues that arise in producing an efficient C implementation of the resulting algorithm. Along the way, we develop some theorems about program refinements, and illustrate the usefulness of the theorems in the context of refining the original Unity program. 8 refs.
- Research Organization:
- Argonne National Lab. (ANL), Argonne, IL (United States)
- Sponsoring Organization:
- USDOD; DOE/ER
- DOE Contract Number:
- W-31109-ENG-38
- OSTI ID:
- 5005530
- Report Number(s):
- CONF-900329-2; ON: DE90005654; CNN: N00014-86-K-0763; N00014-87-K-0510
- Resource Relation:
- Journal Volume: 18; Journal Issue: 6; Conference: ACM SIGPLAN symposium on principles and practice of parallel programming, Seattle, WA (USA), 15-16 Mar 1990
- Country of Publication:
- United States
- Language:
- English
Similar Records
Parallel computation of genetic likelihoods using CRI-MAP, PVM, and a network of distributed workstations
Analysis of the numerical effects of parallelism on a parallel genetic algorithm