Proceedings Article10.1109/HICSS.1995.375471
Automatic task graph generation techniques
Michel Cosnard,Michel Loi +1 more
- 04 Jan 1995
- Vol. 2, pp 113-122
TL;DR: A model of parallel computation, the parameterized task graph, is presented, which is a compact, problem-size-independent representation of some frequently-used directed acylic task graphs, and techniques for automating the construction of such a representation are proposed.
read more
Abstract: Presents a model of parallel computation, the parameterized task graph, which is a compact, problem-size-independent representation of some frequently-used directed acylic task graphs. Techniques for automating the construction of such a representation, starting from an annotated sequential program, are proposed. We show how to represent the computational load for each node and the communication volume for each edge in a problem-size-independent way. Lastly, we describe related work and techniques required in order to generate local data structures, computation code and communication code. >
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
Flexible Development of Dense Linear Algebra Algorithms on Massively Parallel Architectures with DPLASMA
George Bosilca,Aurelien Bouteiller,Anthony Danalis,Mathieu Faverge,Azzam Haidar,Thomas Herault,Jakub Kurzak,Julien Langou,Pierre Lemarinier,Hatem Ltaief,Piotr Luszczek,Asim YarKhan,Jack Dongarra +12 more
- 16 May 2011
TL;DR: It is demonstrated through experimental results on the Cray XT5 Kraken system that the DAG-based approach has the potential to achieve sizable fraction of peak performance which is characteristic of the state-of-the-art distributed numerical software on current and emerging architectures.
Patent
Rule processing system
David L. Huelsman,Sharon E. Love,Douglas M. Mair +2 more
- 27 Feb 2007
TL;DR: A rule processing apparatus as discussed by the authors includes modules for defining/entering attributes, enumerations, and relationships; packaging the definitions in a reduced canonical form suitable for propositional logic manipulation using zero-suppressed binary decision diagrams (Zdd) to produce a prime Zdd; and executing the rule by applying a series of user inputs to the prime ZDD to determine a result that preferably includes conflict and selection advice to guide the user to satisfaction.
80
PTG: an abstraction for unhindered parallelism
Anthony Danalis,George Bosilca,Aurelien Bouteiller,Thomas Herault,Jack Dongarra +4 more
- 16 Nov 2014
TL;DR: The Parameterized Task Graph language decouples the expression of the parallelism in the algorithm from the control-flow ordering, load balance, and data distribution, which means programs are more adaptable and map more efficiently on challenging hardware, as well as maintain portability across diverse architectures.
A resource-aware scheduling algorithm with reduced task duplication on heterogeneous computing systems
TL;DR: A novel duplication-based algorithm to schedule tasks with the least redundant duplications is proposed and an optimizing scheme is introduced to search and remove redundancy for a schedule generated by the proposed algorithm further.
39
Power monitoring with PAPI for extreme scale architectures and dataflow-based programming models
Heike McCraw,James Ralph,Anthony Danalis,Jack Dongarra +3 more
- 01 Dec 2014
TL;DR: Detailed information is provided about three components that allow power monitoring on the Intel Xeon Phi and Blue Gene/Q and the integration of PAPI in PARSEC - a task-based dataflow-driven execution engine - enabling hardware performance counter and power monitoring at true task granularity.
References
•Book
Theory of Linear and Integer Programming
Alexander Schrijver
- 01 Dec 1986
TL;DR: Introduction and Preliminaries.
Some efficient solutions to the affine scheduling problem: I. One-dimensional time
TL;DR: This paper deals with the problem of finding closed form schedules as affine or piecewise affine functions of the iteration vector and presents an algorithm which reduces the scheduling problem to a parametric linear program of small size, which can be readily solved by an efficient algorithm.
654
Supernode partitioning
François Irigoin,R. Triolet +1 more
- 13 Jan 1988
TL;DR: A class of partitionings is presented that encompasses previous techniques and provides enough flexibility to adapt code to multiprocessors with two levels of parallelism and two level of memory.
635
Parametric integer programming
TL;DR: In this paper, the analysis semantique des programs informatiques conduit a la resolution de problemes de programmation parametrique entiere, i.e., the problem of finding a parametrization of a program.
478
(Pen)-ultimate tiling?
TL;DR: In this article, Tiling is used to group computation points so as to increase computation granularity and to reduce the overhead due to communication time, and the authors provide a constructive method for defining the optimal tiling in a scalable environment.
129