In this paper, we examine and characterize effects of communication interactions of parallel and sequential jobs on a nondedicated ATM network of workstations. We quantitatively model the interaction process and measure communication delays of parallel jobs caused by the interaction. In addition, a scheme is proposed to distinguish and prioritize the two types of communications messages to reduce the disturbance of sequential jobs' communications on the communication performance of parallel jobs. Measurement and simulation results on an ATM network of workstations support our analytical models and the scheme. Keywords: Communication, networks of workstations (NOW), parallel processing, TCP/IP. 1 Introduction The wide availabil...
Parallel computers can take many different architectural forms. One cost effective and convenient me...
. Efficient communication in networks is a prerequisite to exploit the performance of large parallel...
This work provides a systematic study of the impact of commu-nication performance on parallel applic...
The objective of this work is to compare the performance of three common environments for supporting...
Cluster-based computing, which exploits the aggregate power of networked collections of computers, h...
This paper examines the plausibility of using a network of workstations (NOW) for a mixture of paral...
. In this paper, we describe experiments comparing the communication times for a number of different...
This paper examines the plausibility of using a network of workstations (NOW) for a mixture of paral...
Recent developments in communication architectures for parallel machines have made significant progr...
We consider parallel computing on a network of workstations using a connection-oriented protocol (e....
In order to be able to develop robust and effective parallel applications and algorithms, one should...
Clusters of workstations are a popular platform for high-performance computing. For many parallel ap...
The present work illustrates the architecture and utilization of a simulated ATM testbed, which enab...
For communication-intensive parallel applications, the maximum degree of concurrency achievable is l...
This paper presents the design and implementation of an efficient communication system, Pupa, devel...
Parallel computers can take many different architectural forms. One cost effective and convenient me...
. Efficient communication in networks is a prerequisite to exploit the performance of large parallel...
This work provides a systematic study of the impact of commu-nication performance on parallel applic...
The objective of this work is to compare the performance of three common environments for supporting...
Cluster-based computing, which exploits the aggregate power of networked collections of computers, h...
This paper examines the plausibility of using a network of workstations (NOW) for a mixture of paral...
. In this paper, we describe experiments comparing the communication times for a number of different...
This paper examines the plausibility of using a network of workstations (NOW) for a mixture of paral...
Recent developments in communication architectures for parallel machines have made significant progr...
We consider parallel computing on a network of workstations using a connection-oriented protocol (e....
In order to be able to develop robust and effective parallel applications and algorithms, one should...
Clusters of workstations are a popular platform for high-performance computing. For many parallel ap...
The present work illustrates the architecture and utilization of a simulated ATM testbed, which enab...
For communication-intensive parallel applications, the maximum degree of concurrency achievable is l...
This paper presents the design and implementation of an efficient communication system, Pupa, devel...
Parallel computers can take many different architectural forms. One cost effective and convenient me...
. Efficient communication in networks is a prerequisite to exploit the performance of large parallel...
This work provides a systematic study of the impact of commu-nication performance on parallel applic...