Proceedings Article10.1109/SHPCC.1994.296705
Randomized load balancing for tree-structured computation
Soumen Chakrabarti,Abhiram Ranade,Katherine Yelick +2 more
- 23 May 1994
- pp 666-673
TL;DR: This work shows that the time taken to explore a task tree is likely to be within a small constant factor of an inherent lower bound for the tree instance, and thus extends earlier work which assumed fixed cost tasks and used a bulk synchronous style.
read more
Abstract: Studies the performance of a randomized algorithm for balancing load across a multiprocessor executing a dynamic irregular task tree. Specifically, we show that the time taken to explore a task tree is likely to be within a small constant factor of an inherent lower bound for the tree instance. Our model permits arbitrary task times and overlap between computation and load balance, and thus extends earlier work (R.M. Karp and Y. Zhang, 1988) which assumed fixed cost tasks and used a bulk synchronous style in which the system alternated between distinct computing and load balancing steps. Our analysis is supported by experiments with application codes, demonstrating that the efficiency is high enough to make this method practical. >
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
On runtime parallel scheduling for processor load balancing
TL;DR: Scheduling algorithms for tree, hypercube, and mesh networks are presented that can fully balance the load and maximize locality at runtime and are significantly reduced compared to other existing algorithms.
66
Dynamic Load Balancing for the Distributed Mining of Molecular Structures
G. Di Fatta,Michael R. Berthold +1 more
TL;DR: The three main aspects of the proposed distributed algorithm are described, namely, a dynamic partitioning of the search space, a distribution process based on a peer-to-peer communication framework, and a novel receiver-initiated load balancing algorithm.
Runtime incremental parallel scheduling (RIPS) on distributed memory computers
Wei Shu,Min-You Wu +1 more
TL;DR: The paper presents methods for scheduling a single job on a dedicated parallel machine using the advanced parallel scheduling technique to produce a low overhead, high quality load balancing, as well as adapting to irregular applications.
35
Multipol: A Distributed Data Structure Library
Soumen Chakrabarti,Etienne Deprit,Eun Kyung Im,Jeffrey A. Jones,Arvind Krishnamurthy,Chi Wen,Katherine Yelick +6 more
- 01 Jul 1995
TL;DR: An overview of Multipol is given, which includes parallel versions of classic data structures such as trees, sets, lists, graphs, and queues, built on a portable runtime layer that provides basic communication, synchronization, and caching.
24
Runtime Support for Portable Distributed Data Structures
Chih-Po Wen,Soumen Chakrabarti,Etienne Deprit,Arvind Krishnamurthy,Katherine Yelick +4 more
- 01 Jan 1996
TL;DR: The Multipol runtime layer is described, which provides an efficient and portable abstraction underlying the data structures, which contains a thread system to express computations with varying degrees of parallelism and to support multiple threads per processor for hiding communication latency.
23
References
LogP: towards a realistic model of parallel computation
David E. Culler,Richard M. Karp,David A. Patterson,Abhijit Sahay,Klaus Erik Schauser,Eunice E. Santos,Ramesh Subramonian,Thorsten von Eicken +7 more
- 01 Jul 1993
TL;DR: A new parallel machine model, called LogP, is offered that reflects the critical technology trends underlying parallel computers and is intended to serve as a basis for developing fast, portable parallel algorithms and to offer guidelines to machine designers.
•Book
Fundamentals of matrix computations
David S. Watkins
- 01 Jan 1991
TL;DR: This paper focuses on Gaussian Elimination as a model for Iterative Methods for Linear Systems, and its applications to Singular Value Decomposition and Sparse Eigenvalue Problems.
1.1K
A dynamic scheduling method for irregular parallel programs
Steven Lucco
- 01 Jul 1992
TL;DR: A fundamental relationship between three quantities that characterize an irregular parallel computation is shown: the total available parallelism, the optimal grain size, and the statistical variance of execution times for individual tasks, which yields a dynamic scheduling algorithm that substantially reduces the overhead of executing irregular parallel operations.
83
Communication complexity for parallel divide-and-conquer
I-Chen Wu,Hsiang-Tsung Kung +1 more
- 01 Sep 1991
TL;DR: A scheduling algorithm is proposed, and lower bounds on the communication cost are derived, since the parallel computation cost of the algorithm is near optimal.