The support on cluster environments of ”legacy proto-cols ” is important to avoid rewriting the code of applica-tions, but this support should not prevent to achieve the maximum communication performance. This paper ad-dresses this issue by implementing the Berkeley sockets in-terface over the MMP message passing protocol, which is the lower layer of the Maestro2 cluster communication net-work. Experimental results show that MMP-Sockets offers a minimum latency of 25s and a maximum throughput of 1250Mbps. These values correspond to a relative increase of 80 % for the latency and a decrease of about 30 % for the throughput, regarding to a communication based only on MMP. However, MMP-Sockets increases the compatibility and portability of dev...
This paper presents an efficient communication subsystem, DP-II, for clustering standard high-volume...
Constructing a system with traditional resources, based on the concept of high-performance computati...
Parallel computing on clusters of workstations and personal computers has very high potential, sinc...
A cluster has become a popular platform for high per-formance computing. Currently, many PC clusters...
Most clusters so far have used WAN or LAN-based net-work products for communication due to their mar...
This work covers the implementation of a raw Ethernet communication module for the Open MPI message ...
This paper describes an Operating System approach to the problem of delivering low latency high band...
This paper describes the basic concepts of our solution to improve the performance of Ethernet Commu...
The Sockets Direct Protocol (SDP) has been proposed recently in order to enable sockets based applic...
There is a need for systems to provide additional processing to extract useful information from the ...
InfiniBand has been recently standardized by the industry to design next generation high-end cluster...
Abstract:- The recent years have seen a considerable increase in the number of cluster systems. Thes...
This paper describes the basic concepts of our solution to improve the performance of Ethernet Commu...
A new trend is emerging to replace massively parallel machines with clusters built from Commercial O...
This paper investigates the use of ATM for cluster based computing. The need for a native ATM API is...
This paper presents an efficient communication subsystem, DP-II, for clustering standard high-volume...
Constructing a system with traditional resources, based on the concept of high-performance computati...
Parallel computing on clusters of workstations and personal computers has very high potential, sinc...
A cluster has become a popular platform for high per-formance computing. Currently, many PC clusters...
Most clusters so far have used WAN or LAN-based net-work products for communication due to their mar...
This work covers the implementation of a raw Ethernet communication module for the Open MPI message ...
This paper describes an Operating System approach to the problem of delivering low latency high band...
This paper describes the basic concepts of our solution to improve the performance of Ethernet Commu...
The Sockets Direct Protocol (SDP) has been proposed recently in order to enable sockets based applic...
There is a need for systems to provide additional processing to extract useful information from the ...
InfiniBand has been recently standardized by the industry to design next generation high-end cluster...
Abstract:- The recent years have seen a considerable increase in the number of cluster systems. Thes...
This paper describes the basic concepts of our solution to improve the performance of Ethernet Commu...
A new trend is emerging to replace massively parallel machines with clusters built from Commercial O...
This paper investigates the use of ATM for cluster based computing. The need for a native ATM API is...
This paper presents an efficient communication subsystem, DP-II, for clustering standard high-volume...
Constructing a system with traditional resources, based on the concept of high-performance computati...
Parallel computing on clusters of workstations and personal computers has very high potential, sinc...