This article presents some performance evaluating results obtained by measuring computer clusters used for high computational capacity. For each cluster, named AIX, Linux, and SP2, measurements was taken comprising the data throughput in the transport layer (TCP and UDP). Using statistical results of the collected data, a linear relation between latency and segment size was determined. From the analysis of the clusters we observed that there is a strong relation between segment size and the latency for collision free environments. Analyzing the clusters, the smaller latency was shown by the Linux cluster with Fast-Ethernet. Although the AIX cluster throughput is greater than the Linux cluster throughput, other factors such as negotiation, e...
Minimum Cycle Time is a common performance indicator adopted to compare Real-Time Ethernet protocols...
The Department of Telecommunications and Signal Processing (ITS) at the Blekinge Institute of Techno...
Building high performance computer systems requires an understanding of the behaviour of systems and...
Cluster interconnect performance is typically characterized by latency and throughput. However, not ...
This work provides a systematic study of the impact of commu-nication performance on parallel applic...
It is essential to know how the system used in a company performs and at what eventual point the sys...
Abstract. Beowulf (PC) clusters represent a cost-effective platform for large scale scientific compu...
This paper addressed the impacts of latency on good throughputs of a computer network. The networks ...
<p>ICPE 2018 published paper artifact submission. Abstract:</p> <p>In this paper, we study the imp...
This is an extension of a similar paper from Workshop on Communication Architecture for Clusters (CA...
A critical factor in the design of any parallel processing computer system is the interconnection ne...
The performance1 of Linux clusters used for High-Performance Computation (HPC) applications is affec...
• Good understanding of the TCP protocols and ability to recompile the linux kernel • Basics of expe...
Networked clusters of computers are commonly used to either process multiple sequential jobs concurr...
In this paper the authors present performance results from several parallel benchmarks and applicati...
Minimum Cycle Time is a common performance indicator adopted to compare Real-Time Ethernet protocols...
The Department of Telecommunications and Signal Processing (ITS) at the Blekinge Institute of Techno...
Building high performance computer systems requires an understanding of the behaviour of systems and...
Cluster interconnect performance is typically characterized by latency and throughput. However, not ...
This work provides a systematic study of the impact of commu-nication performance on parallel applic...
It is essential to know how the system used in a company performs and at what eventual point the sys...
Abstract. Beowulf (PC) clusters represent a cost-effective platform for large scale scientific compu...
This paper addressed the impacts of latency on good throughputs of a computer network. The networks ...
<p>ICPE 2018 published paper artifact submission. Abstract:</p> <p>In this paper, we study the imp...
This is an extension of a similar paper from Workshop on Communication Architecture for Clusters (CA...
A critical factor in the design of any parallel processing computer system is the interconnection ne...
The performance1 of Linux clusters used for High-Performance Computation (HPC) applications is affec...
• Good understanding of the TCP protocols and ability to recompile the linux kernel • Basics of expe...
Networked clusters of computers are commonly used to either process multiple sequential jobs concurr...
In this paper the authors present performance results from several parallel benchmarks and applicati...
Minimum Cycle Time is a common performance indicator adopted to compare Real-Time Ethernet protocols...
The Department of Telecommunications and Signal Processing (ITS) at the Blekinge Institute of Techno...
Building high performance computer systems requires an understanding of the behaviour of systems and...