Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 20 of 24 for “"Myrinet"”.
-
MPI collective operations over Myrinet
… than a host-based implementation over the same Myrinet, and for MPLComm.Create, the improvement factor is 2. The NIC-based MPI_Bcast is always better than a host-based implementation for all message sizes, and for a small message, the improvement factor of the broadcast latency is 2 to 3 times. …
-
Implementation of processor farms on myrinet network interface cards
… sixteen Pentium II computers interconnected by a Myrinet [18] gigabit per second programmable network. We developed and optimized a processor farm, which we call pfarm for our sixteen node cluster. We show that pfarm effectively executes relatively fine grain tasks with variable execution times, …
-
Performance evaluation of Fast Ethernet, ATM and Myrinet under PVM
… Transfer Mode (ATM) OC-3c (155Mb/s) and Myrinet. Together, they represent an interesting range of alternatives for parallel cluster computing. A characterization of network delays and throughput and a comparison of the expected costs of the three environments are developed to provide a …
-
Τεχνικές εισόδου-εξόδου και χρονοδρομολόγησης για την αποδοτική χρήση μοιραζόμενων αρχιτεκτονικών πόρων σε συστοιχίες κόμβων συμμετρικής πολυεπεξεργασίας
… performance interconnection networks, such as Myrinet. The increasing use of clusters for data-intensive workloads, in combination with the trend for ever-increasing cores per processor die, poses significant load on the I/O subsystem. Thus, its performance becomes decisive in determining …
-
Using nodes in a cluster efficiently
… Our performance results show that the Opteron/Myrinet cluster is able to achieve significantly better utilization of the second processor than the Xeon/Myrinet cluster.</p>
-
Hierarchical Simulation to Assess Hardware and Software Dependability
… commercial, high-speed networking system called Myrinet. One key result from the case studies shows that the simulation method predicts that same fault impact 87.5% of the time, as is obtained by similar fault injections into a real Myrinet system. Reasons for the remaining discrepancy are …
-
A network processor based message manager for MPI
… is an MPI specific messaging system for the Myrinet System Area Networks (SAN). It consists of a lowlevel message manager executing on the LANai processor of the Myrinet Network Interface Card (NIC), a thin host interface layer, and LAM-MPI, a public domain version of MPI. MPI-NP II is a …
-
Αποδοτικός διαμοιρασμός πόρων εισόδου / εξόδου σε περιβάλλοντα εικονικών μηχανών
… framework, binary compatible with Myrinet/MX and wire compatible with MXoE. Its design is based on the Open-MX protocol, a port of the Myrinet/MX over generic Ethernet adapters. Xen2MX combines the zero-copy characteristics of Open-MX with Xen's memory sharing techniques; the …
-
The MPI implementation on NetVM
… and real-world applications compared to LAM over Myrinet. The remote memory model is shown to be powerful in implementing MPI semantics.
-
Implementação da biblioteca de comunicação DECK sobre o padrão de protocolo de comunicação em nível de usuário VIA
… a implementação VI-GM de VIA para redes Myrinet. A biblioteca DECK/VIA demonstrou uma latência de 86.85 μs e uma largura de banda máxima de 205 Mbytes/s, 82% da banda nominal da rede Myrinet. Para validar a biblioteca foi executada a aplicação FT do pacote NPB. Apresenta-se comparações …
-
An MPI messaging layer for network processors
… the design and implementation of MPI-NP, a Myrinet communication system tailored to support LAM, a public domain version of MPI. The goals of MPI-NP are to reduce the time spent on the host for message processing, and to promote the overlap of computation and communication. MPI-NP achieves …
-
Communication performance measurement and analysis on commodity clusters.
… memory machine. A comparison was done between Myrinet and Ethernet network performance on the same machine, an IBM Linux cluster with 128 dual processor nodes, using the MPICH MPI library. The analysis focused mainly on the scalability and variability of communication times for the different …
-
Adaptive turn-prohibition routing algorithm for the networks of workstations
… schemes, like up/down routing algorithm used in Myrinet switches. However, most of existing approaches for deadlock-free routing either try to eliminate any possibility of deadlock occurrence, which can result in putting extra restrictions on the routing in the networks or put no restrictions on …
-
Evaluating communication performance in large parallel computers
… packing on a SGI Origin 2000, a Cray T3E-900, a Myrinet Linux cluster and an Ethernet Linux cluster. Four communication tests using MPI derived type routines and packing/unpacking techniques are run for a variety of message sizes using four processors on these machines. Except for one test, MPI …
-
Uma Biblioteca para programação paralela por troca de mensagens de clusters baseados na tecnologia SCI
… consecução de alto desempenho sobre a tecnologia Myrinet. Os resultados auferidos com o uso do DECK/SCI revelam a eficiência dos mecanismos projetados, e a utilização profícua das características de alto desempenho intrínsecas da rede SCI, haja visto que se obteve uma performance muito próxima dos …
-
Lunar: A User-Level Stack Library for Network Emulation
… switched Ethernet) to approximately 39% (1 Gbps Myrinet).
-
Improving Message-Passing Performance and Scalability in High-Performance Clusters
… interconnects: InfiniBand, iWARP Ethernet, and Myrinet. Using novel techniques, I assess characteristics such as overlap and communication progress ability, buffer reuse effect on latency, and multiple-connection scalability. The outcome highlights some of the inefficiencies that exist in the …
-
A Distributed Open Environment for Real-Time Applications
… open system for different types of networks: Myrinet, Controller Area Network (CAN), and Wireless LAN. While the hierarchical scheduling framework remains the same, several scheduling algorithms are developed to address the design issues arise from the hardware constraints and/or industrial …
-
Performance Modelling of Message-Passing Parallel Programs
… namely those provided by Fast Ethernet, Myrinet and QsNet. Network contention, a problem ignored by most research in this area, was found to cause extensive performance variation during message-passing operations. For point-to-point communication, this variation was best described by …
-
Sistema Aldeia : programação paralela e distribuída em Java sobre Infiniband e DECK
… comunicação com uma interface Java sobre redes Myrinet, SCI, além de TCP/IP. Somada a essa vantagem, a utilização do DECK também proporciona a propriedade de geração de rastros para a depuração de programas paralelos escritos com o Aldeia. Uma das grandes vantagens do Aldeia está na sua …
Page 1 of 2