The Experts below are selected from a list of 75 Experts worldwide ranked by ideXlab platform
Yutaka Ishikawa - One of the best experts on this subject based on the ideXlab platform.
-
The design and implementation of an asynchronous communication mechanism for the MPI communication model
2004Co-Authors: Motohiko Matsuda, Tomohiro Kudoh, H. Tazuka, Yutaka IshikawaAbstract:Many implementations of an MPI communication library are realized on top of the socket interface which is based on connection-oriented stream communication. This paper addresses a mismatch between the MPI communication model and the socket interface. In order to overcome a mismatch and implement an efficient MPI library for large-scale commodity-based clusters, a new communication mechanism, called O2G, is designed and implemented. O2G integrates receive queue management of MPI into a TCP/IP Protocol Handler, without modifying the Protocol stacks. Received data is extracted from the TCP receive buffer and copied into the user space within the TCP/IP Protocol Handler invoked by interrupts. It totally avoids polling of sockets and reduces system call overhead, which becomes dominant in large-scale clusters. In addition, its immediate and asynchronous receive operation avoids message flow disruption due to a shortage of capacity in the receive buffer, and keeps the bandwidth high. An evaluation using the NAS Parallel Benchmarks shows that O2G made an MPI implementation up to 30 percent faster than the original one. An evaluation on bandwidth also shows that O2G made an MPI implementation independent of the number of connections, while an implementation with sockets was greatly affected by the number of connections.
-
CLUSTER - The design and implementation of an asynchronous communication mechanism for the MPI communication model
2004 IEEE International Conference on Cluster Computing (IEEE Cat. No.04EX935), 1Co-Authors: Motohiko Matsuda, Tomohiro Kudoh, H. Tazuka, Yutaka IshikawaAbstract:Many implementations of an MPI communication library are realized on top of the socket interface which is based on connection-oriented stream communication. This work addresses a mismatch between the MPI communication model and the socket interface. In order to overcome a mismatch and implement an efficient MPI library for large-scale commodity-based clusters, a new communication mechanism, called 02G, is designed and implemented. O2G integrates receive queue management of MPI into a TCP/IP Protocol Handler, without modifying the Protocol stacks. Received data is extracted from the TCP receive buffer and copied into the user space within the TCP/IP Protocol Handler invoked by interrupts. It totally avoids polling of sockets and reduces system call overhead, which becomes dominant in large-scale clusters. In addition, its immediate and asynchronous receive operation avoids message flow disruption due to a shortage of capacity in the receive buffer, and keeps the bandwidth high. An evaluation using the NAS Parallel Benchmarks shows that 02G made an MPI implementation up to 30 percent faster than the original one. An evaluation on bandwidth also shows that 02G made an MPI implementation independent of the number of connections, while an implementation with sockets was greatly affected by the number of connections.
Ekrem Serin - One of the best experts on this subject based on the ideXlab platform.
-
4. TITLE AND SUBTITLE: Design and Test of The Cross-Format Schema For Networked Virtual Environments
2003Co-Authors: Co Advisor, Joseph Sullivan, Second Reader, Curt Blais, Ekrem SerinAbstract:A Networked Virtual Environment (Net-VE) is a distributed software system in which multiple users interact with each other in real time even though these users may be located around the world [Zyda 99]. Net-VEs gained first attention through a variety of DOD and Academic research projects. After release of the multiplayer game DOOM, the gaming industry captured the idea of interactive multiplayer games. Today there are many popular Internet-based multiplayer games available. Effective networking of diverse entities and systems is a common problem for Networked Virtual Environments. In order to communicate with other entities a variety of communication Protocols are used. Historically these communication Protocols are "hard coded" into the software system and all nodes that participate in the environment must identically implement the Protocols to interact with others. These communication Protocols require authoring and compiling by a trained programmer. When the compiling process is introduced to the networked virtual environment, it detracts the extensibility and dynamicism of the system. This thesis presents the design and development of a Networked Virtual Environment model that uses Cross Format Schema Protocol (XFSP). With this work we show that a networked simulation can work for 24 hours a day and 7 days a week with an extensible schema based networking Protocol and it is not necessary to hard code and compile the Protocols into the networked virtual environments. Furthermore, this thesis presents a general automatic Protocol Handler for schema-defined XML document or message. Additionally, this work concludes with idea that Protocols can be loaded and extended at runtime, and can be created with different-fidelity resolutions, resulting in swapping at runtime ba..
Motohiko Matsuda - One of the best experts on this subject based on the ideXlab platform.
-
The design and implementation of an asynchronous communication mechanism for the MPI communication model
2004Co-Authors: Motohiko Matsuda, Tomohiro Kudoh, H. Tazuka, Yutaka IshikawaAbstract:Many implementations of an MPI communication library are realized on top of the socket interface which is based on connection-oriented stream communication. This paper addresses a mismatch between the MPI communication model and the socket interface. In order to overcome a mismatch and implement an efficient MPI library for large-scale commodity-based clusters, a new communication mechanism, called O2G, is designed and implemented. O2G integrates receive queue management of MPI into a TCP/IP Protocol Handler, without modifying the Protocol stacks. Received data is extracted from the TCP receive buffer and copied into the user space within the TCP/IP Protocol Handler invoked by interrupts. It totally avoids polling of sockets and reduces system call overhead, which becomes dominant in large-scale clusters. In addition, its immediate and asynchronous receive operation avoids message flow disruption due to a shortage of capacity in the receive buffer, and keeps the bandwidth high. An evaluation using the NAS Parallel Benchmarks shows that O2G made an MPI implementation up to 30 percent faster than the original one. An evaluation on bandwidth also shows that O2G made an MPI implementation independent of the number of connections, while an implementation with sockets was greatly affected by the number of connections.
-
CLUSTER - The design and implementation of an asynchronous communication mechanism for the MPI communication model
2004 IEEE International Conference on Cluster Computing (IEEE Cat. No.04EX935), 1Co-Authors: Motohiko Matsuda, Tomohiro Kudoh, H. Tazuka, Yutaka IshikawaAbstract:Many implementations of an MPI communication library are realized on top of the socket interface which is based on connection-oriented stream communication. This work addresses a mismatch between the MPI communication model and the socket interface. In order to overcome a mismatch and implement an efficient MPI library for large-scale commodity-based clusters, a new communication mechanism, called 02G, is designed and implemented. O2G integrates receive queue management of MPI into a TCP/IP Protocol Handler, without modifying the Protocol stacks. Received data is extracted from the TCP receive buffer and copied into the user space within the TCP/IP Protocol Handler invoked by interrupts. It totally avoids polling of sockets and reduces system call overhead, which becomes dominant in large-scale clusters. In addition, its immediate and asynchronous receive operation avoids message flow disruption due to a shortage of capacity in the receive buffer, and keeps the bandwidth high. An evaluation using the NAS Parallel Benchmarks shows that 02G made an MPI implementation up to 30 percent faster than the original one. An evaluation on bandwidth also shows that 02G made an MPI implementation independent of the number of connections, while an implementation with sockets was greatly affected by the number of connections.
Yoshiaki Nagaya - One of the best experts on this subject based on the ideXlab platform.
-
development of high performance and flexible Protocol Handler for international web accesses
High Performance Computing and Communications, 2019Co-Authors: Ken T Murata, Praphan Pavarangkoon, Kenji Inoue, Takamichi Mizuhara, Yuya Kagebayashi, Kazunori Yamamoto, Kazuya Muranaga, Eizen Kimura, Yoshiaki NagayaAbstract:In this big data era, a high-speed web access technique is expected for sharing a variety of large-scale data, especially over an international network. There are several high-speed web access techniques, such as data caching, data compression, data visualization, and intelligent data processing. In addition, hypertext transfer Protocol (HTTP)/1.1 provides a persistent connection for better throughput than the HTTP/1.0. However, using these techniques developed so far, there are many users who are not satisfied with the current web access speed. In this paper, we propose a novel Protocol Handler for international web accesses based on a high-speed data transfer Protocol, namely high-performance and flexible Protocol (HpFP). We first design our Protocol Handler on the basis of HTTP, and then implement it for the Internet Explorer browser in Windows operating system (OS). Using a laboratory experiment, we evaluate the performance of our Protocol Handler in terms of throughput on a real-time web for weather satellite data. The results show that the proposed Protocol Handler achieves higher throughput than the HTTP for website requests on international networks. The improvement gets more significant as either the latency or the packet loss ratio becomes higher. This implies that the proposed Protocol Handler especially devotes to international web accesses for big data.
-
HPCC/SmartCity/DSS - Development of High-Performance and Flexible Protocol Handler for International Web Accesses
2019 IEEE 21st International Conference on High Performance Computing and Communications; IEEE 17th International Conference on Smart City; IEEE 5th I, 2019Co-Authors: Ken T Murata, Praphan Pavarangkoon, Kenji Inoue, Takamichi Mizuhara, Yuya Kagebayashi, Kazunori Yamamoto, Kazuya Muranaga, Eizen Kimura, Yoshiaki NagayaAbstract:In this big data era, a high-speed web access technique is expected for sharing a variety of large-scale data, especially over an international network. There are several high-speed web access techniques, such as data caching, data compression, data visualization, and intelligent data processing. In addition, hypertext transfer Protocol (HTTP)/1.1 provides a persistent connection for better throughput than the HTTP/1.0. However, using these techniques developed so far, there are many users who are not satisfied with the current web access speed. In this paper, we propose a novel Protocol Handler for international web accesses based on a high-speed data transfer Protocol, namely high-performance and flexible Protocol (HpFP). We first design our Protocol Handler on the basis of HTTP, and then implement it for the Internet Explorer browser in Windows operating system (OS). Using a laboratory experiment, we evaluate the performance of our Protocol Handler in terms of throughput on a real-time web for weather satellite data. The results show that the proposed Protocol Handler achieves higher throughput than the HTTP for website requests on international networks. The improvement gets more significant as either the latency or the packet loss ratio becomes higher. This implies that the proposed Protocol Handler especially devotes to international web accesses for big data.
Tomohiro Kudoh - One of the best experts on this subject based on the ideXlab platform.
-
The design and implementation of an asynchronous communication mechanism for the MPI communication model
2004Co-Authors: Motohiko Matsuda, Tomohiro Kudoh, H. Tazuka, Yutaka IshikawaAbstract:Many implementations of an MPI communication library are realized on top of the socket interface which is based on connection-oriented stream communication. This paper addresses a mismatch between the MPI communication model and the socket interface. In order to overcome a mismatch and implement an efficient MPI library for large-scale commodity-based clusters, a new communication mechanism, called O2G, is designed and implemented. O2G integrates receive queue management of MPI into a TCP/IP Protocol Handler, without modifying the Protocol stacks. Received data is extracted from the TCP receive buffer and copied into the user space within the TCP/IP Protocol Handler invoked by interrupts. It totally avoids polling of sockets and reduces system call overhead, which becomes dominant in large-scale clusters. In addition, its immediate and asynchronous receive operation avoids message flow disruption due to a shortage of capacity in the receive buffer, and keeps the bandwidth high. An evaluation using the NAS Parallel Benchmarks shows that O2G made an MPI implementation up to 30 percent faster than the original one. An evaluation on bandwidth also shows that O2G made an MPI implementation independent of the number of connections, while an implementation with sockets was greatly affected by the number of connections.
-
CLUSTER - The design and implementation of an asynchronous communication mechanism for the MPI communication model
2004 IEEE International Conference on Cluster Computing (IEEE Cat. No.04EX935), 1Co-Authors: Motohiko Matsuda, Tomohiro Kudoh, H. Tazuka, Yutaka IshikawaAbstract:Many implementations of an MPI communication library are realized on top of the socket interface which is based on connection-oriented stream communication. This work addresses a mismatch between the MPI communication model and the socket interface. In order to overcome a mismatch and implement an efficient MPI library for large-scale commodity-based clusters, a new communication mechanism, called 02G, is designed and implemented. O2G integrates receive queue management of MPI into a TCP/IP Protocol Handler, without modifying the Protocol stacks. Received data is extracted from the TCP receive buffer and copied into the user space within the TCP/IP Protocol Handler invoked by interrupts. It totally avoids polling of sockets and reduces system call overhead, which becomes dominant in large-scale clusters. In addition, its immediate and asynchronous receive operation avoids message flow disruption due to a shortage of capacity in the receive buffer, and keeps the bandwidth high. An evaluation using the NAS Parallel Benchmarks shows that 02G made an MPI implementation up to 30 percent faster than the original one. An evaluation on bandwidth also shows that 02G made an MPI implementation independent of the number of connections, while an implementation with sockets was greatly affected by the number of connections.
