Affiliation:
1. Univ. of Edinburgh, Scotland
Abstract
The Parallel Random Access Machine (PRAM) is an abstract model of parallel computation which allows researchers to focus on the essential characteristics of a parallel architecture and ignore other details. The PRAM has long been acknowledged to be a useful tool for the study of parallel computing, but unfortunately it is not physically implementable in hardware. In order to take advantage of the broad base of algorithms and results regarding this high-level abstraction one needs general methods for allowing the execution of PRAM algorithms on more realistic machines. In the following we survey these methods, which we refer to as PRAM simulation techniques. The general issues of memory management and routing are discussed, and both randomized and deterministic solutions are considered. We show that good theoretical solutions to many of the subproblems in PRAM simulation have been developed, though questions still exist as to their practical utility. This article should allow those performing research in this field to become well acquainted with the current state of the art, while allowing the novice to get an intuitive feeling for the fundamental questions being considered. The introduction should provide a concise tutorial for those unfamiliar with the problem of PRAM simulation.
Publisher
Association for Computing Machinery (ACM)
Subject
General Computer Science,Theoretical Computer Science
Cited by
19 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献
1. Efficient Distributed Algorithms in the k-machine model via PRAM Simulations;2021 IEEE International Parallel and Distributed Processing Symposium (IPDPS);2021-05
2. Connecting MapReduce Computations to Realistic Machine Models;2020 IEEE International Conference on Big Data (Big Data);2020-12-10
3. Graph algorithms: parallelization and scalability;Science China Information Sciences;2020-09-21
4. Substream-Centric Maximum Matchings on FPGA;ACM Transactions on Reconfigurable Technology and Systems;2020-06-10
5. Parallelizing Sequential Graph Computations;ACM Transactions on Database Systems;2018-12-16