Multicomputers: message-passing concurrent computers

Athas, W.C.; Seitz, Charles L.

doi:10.1109/2.73

Cited by 432 publications

(91 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The 3-DSPHB algorithm splits and sorts D into two subsets D U where contains every node in N U that has a higher label than that of the source node and D L where contains every node in N L that has a lower label than that of the source node. 27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59, 60, 61, 62, 63} D L = {24, 23,22,21,20,19,18,17,16,15,14,13,12,11,10,9,8,7,6,5,4 …”

Section: Comparative Studymentioning

confidence: 99%

Three-Dimension Hamiltonian Broadcast Wormhole-Routing

El-Obaid¹

2015

IJCNC

View full text Add to dashboard Cite

Broadcast is one of the most important approach in distributed memory parallel computers that is used to find a routing approach from a one source to all nodes in the mesh. Broadcasting is a data communication task in which corresponds 1.INTRODUCTIONOne of the important and popular architecture in both of scientific computation and high speed computing applications is multicomputer architecture where it is used efficiently in weather forecast and high performance servers. Multicomputer network consists of hundreds or thousands of nodes connected in some fixed topology; each of them contains a microprocessor, local memory, and other supporting devices. Most of the popular direct network topologies fall in the general category of either n-dimensional meshes or k-ary n-cubes because their regular topologies and simple routing. Some sample topologies are shown in Fig. 1. Among topologies, the most commonly used are meshes, tori, hypercubes and trees. The Mesh-based topologies are the most regular simple architecture that is used in multicomputers. This is because that the implementation of mesh is very simple and easy to understand. These properties which are necessary for every topology design [1]. Recent interest in multicomputer systems is therefore concentrated on two-dimensional or three-dimensional mesh and torus networks. Such technology has been adopted by the Intel Touchstone DELTA

show abstract

Section: Comparative Studymentioning

confidence: 99%

Three-Dimension Hamiltonian Broadcast Wormhole-Routing

El-Obaid¹

2015

IJCNC

View full text Add to dashboard Cite

show abstract

“…The reactive kernel and Cantor [4] were other relevant contemporary systems. However, the chare was clearly very similar to the notion of a concurrent object or an actor developed earlier by Agha [2] and Yonezawa [49] et al, which built upon Hewitt's earlier work [17].…”

Section: Introductionmentioning

confidence: 99%

Controlling Concurrency and Expressing Synchronization in Charm++ Programs

Kalé

Lifflander

2014

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Charm++ is a parallel programming system that evolved over the past 20 years to become a well-established system for programming parallel science and engineering applications, in addition to the combinatorial search applications with which it started. At its earliest point, the precursor to Charm++, the Chare Kernel, was a purely reactive specification, similar to most actor languages. This paper describes the evolution of a series of concurrency control mechanisms that have been deployed in Charm++ to tame this unrestricted concurrency in order to improve code clarity and/or to improve performance.

show abstract

“…These expose dependencies at the design level and permit diagrammatic representations such as Figure 12. While in the past message-passing concurrency mapped processes one-to-one to processors [21,42], process-oriented designs are intended to be architecture independent [9,17,25,32].…”

Section: Introductionmentioning

confidence: 99%

Multicore Scheduling for Lightweight Communicating Processes

Ritson

Sampson

Barnes

2009

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Process-oriented programming is a design methodology in which software applications are constructed from communicating concurrent processes. A process-oriented design is typically composed of a large number of small isolated concurrent components. These components allow for the scalable parallel execution of the resulting application on both shared-memory and distributed-memory architectures. In this paper we present a runtime designed to support process-oriented programming by providing lightweight processes and communication primitives. Our runtime scheduler, implemented using lock-free algorithms, automatically executes concurrent components in parallel on multicore systems. Runtime heuristics dynamically group processes into cache-affine work units based on communication patterns. Work units are then distributed via wait-free work-stealing. Initial performance analysis shows that, using the algorithms presented in this paper, process-oriented software can execute with an efficiency approaching that of optimised sequential and coarse-grain threaded designs.

show abstract

Multicomputers: message-passing concurrent computers

Cited by 432 publications

References 5 publications

Three-Dimension Hamiltonian Broadcast Wormhole-Routing

Three-Dimension Hamiltonian Broadcast Wormhole-Routing

Controlling Concurrency and Expressing Synchronization in Charm++ Programs

Multicore Scheduling for Lightweight Communicating Processes

Contact Info

Product

Resources

About