OpenMP on Distributed Memory Computers with FDSM Distributed Shared Memory System (2008)
Hiroya Matsuba, Yutaka Ishikawa
FDSM, a software distributed shared memory system, has been designed and implemented to run OpenMP programs on distributed memory computers. FDSM inspects the access pattern of an application at the...
The RWC PC Cluster II, consisting of 128 Intel (2008)
Yutaka Ishikawa, Hiroshi Tezuka, Atsushi Hori, Shinji Sumimoto, Toshiyuki Takahashi, Hiroshi Harada
Atsushi Hori, Yutaka Ishikawa, Hiroki Konaka, Munenori Maeda, Takashi Tomokiyo
We propose a new process scheduling queue system called the Distributed Queue Tree (DQT) for a distributed memory, dynamically partitionable paral-lel machines. We assume that partitions can be...
Effects of Packet Pacing for MPI Programs in a Grid Environment (2008)
Ryousei Takano, Motohiko Matsuda, Tomohiro Kudoh, Yuetsu Kodama, Fumihiro Okazaki, Yutaka Ishikawa, ...
Abstract — Improving the performance of TCP communication is the key to the successful deployment of MPI programs in a Grid environment in which multiple clusters are connected through high...
Hiroya Matsuba, Yutaka Ishikawa
Operating a cluster on a single IP address is required when the cluster is used to provide some Internet services. This paper proposes SAPS, a new method to assign a single IP address to a cluster....
Network Performance Model for TCP/IP Based Cluster Computing (2008)
Akihiro Nomura, Hiroya Matsuba, Yutaka Ishikawa
Abstract — A new communication model, called the PlogPT model, is proposed to predict communication performance in a commodity cluster where computing nodes communicate using TCP/IP. This model...
Workstation Cluster Software Environment -- MPC++ Approach -- (2008)
Yutaka Ishikawa, Atsushi Hori, Hiroshi Tezuka, Mitsuhisa Sato
e implemented. Advantages of MPC ++ approach to the workstation cluster software environment are programmability and portability. SCore-D daemon processes on workstations are communicated with each...
The MPC++ Multi-Thread Template Library on MPI Document Revision 0.1 (2007)
Abstract This document describes a C++ template library for multi-threaded programming in MPC ++, called MPC ++ MultiThread Template library, and its implementation using the MPI (Message Passing...
Parallel Programming With Distributed Object Groups - The TACO Approach (2007)
Jörg Nolte, J Org Nolte, Mitsuhisa Sato, YUTAKA ISHIKAWA
This paper introduces Taco's basic abstractions
TACO - Dynamic Distributed Collections with Templates and Topologies (2007)
Jörg Nolte, Mitsuhisa Sato, Yutaka Ishikawa
. High-level data-parallel programming with distributed object sets eases many aspects of parallel programming. In this paper we describe the design and implementation of Taco, a template library...
Toward A Seamless Computing Environment -- (2007)
Next Generation Of, Yutaka Ishikawa, Mitsuhisa Sato, Junichi Shimada
Introduction High-speed networks such as Fibre Channel[1], Memory Channel[4], SCI[6], ServerNet, and Myrinet[2] are starting to be used in building parallel systems. These networks achieve a...
Jens Gerlach, Mitsuhisa Sato, Yutaka Ishikawa
. In this paper we discuss design and implementation of a parallel particle simulation method from the realm of material sciences (laser ablation). To achieve a high degree of portability and...
OBP Lib: An Object-Oriented Parallel Library and its Preliminary Performance (2007)
Motohiko Matsuda, Mitsuhisa Sato, Yutaka Ishikawa
OBP Lib (Object-Oriented Building Blocks in Parallel Programming) is a collection of parallel object classes including parallel array and matrix/vector classes in C+ + , which is useful for combining...
Ocore and VAST - Integrating Persistence into a Parallel Object-Oriented Language (2007)
Jörg Nolte, Hiroki Konaka, Atsushi Hori, Yutaka Ishikawa, Munenori Maeda, Takashi Tomokiyo
. State of the art parallel object oriented languages allow the construction of rather complex parallel and distributed data structures. The handling of such data structures on secondary storage for...
Communication Performance of Gigabit LAN Workstation Cluster RWC/WSC (2007)
Kazuto Kubota, Yoshio Tanaka, Motohiko Matsuda, Mitsuhisa Sato, Hiroshi Tezuka, Atsushi Hori, ...
This paper reports on the communication performance of the RWC/WSC. It is then compared with massively parallel machines: the CM5 and the Paragon. For comparison, specific collective communications...
Implementing MPI in a High-Performance, Multithreaded Language MPC (2007)
Francis B. O'Carroll, Atsushi Hori, Hiroshi Tezuka, Yutaka Ishikawa, Satoshi Matsuoka
Device Interface (ADI) Device Dependent Internals Channel Interface Hardware MPID_ControlMsgAvail MPID_RecvAnyControl MPID_SendControl MPID_SendChannel MPID_RecvChannel Fig. 1 MPICH software...
Jorg Nolte, Mitsuhisa Sato, Yutaka Ishikawa
Collective operations on distributed data sets promote a high-level data-parallel programming style that eases many aspects of parallel programming on distributed computing platforms. In this paper...
Jorg Nolte, Mitsuhisa Sato, Yutaka Ishikawa
Collective operations on distributed data sets foster a high-level data-parallel programming style that eases many aspects of parallel programming significantly. In this paper we describe how...
Abstract. A preemptive gang scheduler is developed and evaluated. The gang scheduler, called SCore-D, is implemented on top of a UNIX operating system and runs on workstation and PC clusters...
Jorg Nolte, Mitsuhisa Sato, Yutaka Ishikawa
TACO (Topologies and Collections) is a template library that introduces the flavour of distributed data parallel processing by means of reusable topology classes and C++ templates. This paper...
RWCP PC Cluster Programming Environment-- Extended Abstract-- (2007)
Yutaka Ishikawa, Atsushi Hori, Hiroshi Tezuka, Shinji Sumimoto, Toshiyuki Takahashi, Hiroshi Harada
Wehave been developing the SCore cluster system software running on top of Linux. As shown in Figure 1, the SCore System software consists of a global
The Design and Evaluation of High Performance Communication using a Gigabit Ethernet (2007)
Toshiyuki Takahashi, Yutaka Ishikawa
A high performance communication facility, called the GigaE PM, has been designed and implemented for parallel applications on clusters of computers using a Gigabit Ethernet. The GigaE PM provides...
Real World Computing Partnership (2007)
Yutaka Ishikawa, Atsushi Hori, Hiroshi Tezuka, Shinji Sumimoto, Toshiyuki Takahashi, Hiroshi Harada
A parallel programming system, called MPC++, provides parallel primitives such as a remote function invocation, a global pointer, and a synchronization structure using the C++ template feature. The...
RWCP PC Cluster Programming Environment-- Extended Abstract-- (2007)
Yutaka Ishikawa, Atsushi Hori, Hiroshi Tezuka, Shinji Sumimoto, Toshiyuki Takahashi, Hiroshi Harada
We have been developing the SCore cluster system software running on top of Linux. As shown in Figure 1, the SCore System software consists of a global
Real World Computing Partnership (2007)
Toshiyuki Takahashi, Hiroshi Tezuka, Atsushi Hori, Hiroshi Harada, Yutaka Ishikawa, Peter H. Beckmanz
Abstract. An MPI library, called MPICH-PM/CLUMP, has been implemented on a cluster of SMPs. MPICH-PM/CLUMP realizes zero copy message passing between nodes while using one copy message passing within...
Real World Computing Partnership (2007)
Yutaka Ishikawa, Hiroshi Tezuka, Atsushi Hori, Shinji Sumimoto, Toshiyuki Takahashi, Hiroshi Harada
The RWC PC Cluster II, consisting of 128 Intel Pentium Pro microprocessors connected by a Myricom Myrinet giga-bit network, achieves the comparable speed of a super computer such as Cray T3E. Its...
Proportional assist ventilation using a disturbance observer and predictive control (2006)
Ozaki, Kenji, Soga, Kazutoshi, Ishikawa, Yutaka, Shin, Seiichi, Marukawa, Seishiro, Yamauchi, Junko, ...
Design and evaluation of precise software pacing mechanisms for fast long-distance networks (2005)
Ryousei Takano, Tomohiro Kudoh, Yuetsu Kodama, Motohiko Matsuda, Hiroshi Tezuka, Yutaka Ishikawa
Abstract — In this paper, we propose precise software pacing mechanisms at end nodes of communication paths. First, we propose an Inter Packet Gap (IPG) control mechanism, which inserts gap packets...
Motohiko Matsuda, Tomohiro Kudoh, Hiroshi Tazuka, Yutaka Ishikawa
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...
OpenMP compiler for a Software Distributed Shared Memory System SCASH (Extended Abstract) (2000)
Mitsuhisa Sato, Hiroshi Harada, Yutaka Ishikawa
In this paper, we present an implementation of OpenMP compiler for a page-based software distributed shared memory system, SCASH on a cluster of PCs. For programming distributed memory...
MPICH-PM: Design and implementation of zero copy MPI for PM (1998)
Hiroshi Tezuka, Atsushi Hori, Yutaka Ishikawa
This report describes the design and implementation of a high performance MPI library using a zero copy message transfer primitive supported by PM. MPICH-PM consists of the MPICH implementation of...
Pin-down Cache: A Virtual Memory Management Technique for Zero-copy Communication (1998)
Hiroshi Tezuka, Francis O'Carroll, Atsushi Hori, Yutaka Ishikawa
The overhead of copying data through the central processor by a message passing protocol limits data transfer bandwidth. If the network interface directly transfers the user's memory to the...
Pin-down Cache: A Virtual Memory Management Technique for Zero-copy Communication (1998)
Hiroshi Tezuka, Francis O'Carroll, Atsushi Hori, Yutaka Ishikawa
We propose a zero copy message transfer with a pin-down cache technique which reuses the pinned-down area to decrease the number of calls to pin-down and release primitives. The proposed facility has...
High Performance Communication using a Gigabit Ethernet (1998)
Shinji Sumimoto, Hiroshi Tezuka, Atsushi Hori, Hiroshi Harada, Toshiyuki Takahashi, Yutaka Ishikawa
A high performance communication facility, called the GigaE PM, has been designed and implemented for parallel applications on clusters of computers using a Gigabit Ethernet. The GigaE PM provides...
OBP Lib: An Object-Oriented Parallel Library and its Preliminary Performance (1998)
Motohiko Matsuda, Mitsuhisa Sato, Yutaka Ishikawa
OBP Lib (Object-Oriented Building Blocks in Parallel Programming) is a collection of parallel object classes including parallel array and matrix/vector classes in C+ + , which is useful for combining...
Janus - a C++ Template Library for Parallel Dynamic Mesh Applications (1998)
Jens Gerlach, Mitsuhisa Sato, Yutaka Ishikawa
. We propose Janus --- a C++ template library of container classes and communication primitives for parallel dynamic mesh applications. The paper focuses on two phase containers that are a central...
PM: An Operating System Coordinated High Performance Communication Library (1997)
Hiroshi Tezuka, Atsushi Hori, Yutaka Ishikawa, Mitsuhisa Sato
. We have developed a new communication library, called PM, for the Myrinet gigabit LAN card, that has a dedicated processor and on-board memory to handle a communication protocol. In order to obtain...
Efficient Implementation of Portable C*-like Data-Parallel Library in C++ (1997)
Motohiko Matsuda, Mitsuhisa Sato, Yutaka Ishikawa
The C* language is a data-parallel extension of the C language which incorporates parallel data types. Since the C+ + language provides operator overloading, a C+ + library can implement the C*...
Jens Gerlach, Mitsuhisa Sato, Yutaka Ishikawa
. Finite element meshes are large, richly structured sets whose internal relationships must be visible to different parts of a finite element program. This causes software engineerings problems that...
Parallel Array Class Implementation Using C++ STL Adaptors (1997)
Motohiko Matsuda, Mitsuhisa Sato, Yutaka Ishikawa
. STL Adaptors can combine operations and are used in elimination of temporaries in a C+ + array class; this technique is known as Expression Templates or Template Closures. Since the technique is...
User-level Parallel Operating System for Clustered Commodity Computers (1997)
Atsushi Hori Hiroshi, Hiroshi Tezuka, Yutaka Ishikawa
this paper, after introducing an overview of the design and implementation of SCore-D, the overhead of the SCore-D user-level operating system is evaluated to examine if our approach is a right way...
Parallel Array Class Implementation Using C++ STL Adaptors (Extended Abstract) (1997)
Motohiko Matsuda, Mitsuhisa Sato, Yutaka Ishikawa
this paper, we present an implementation of a parallel array class with ghost-cells employing the Adaptor technique. Ghost-cells, also called as guard-cells or an envelope, are overlapping boundary...
Global State Detection using Network Preemption (1997)
Atsushi Hori Hiroshi, Hiroshi Tezuka, Yutaka Ishikawa
Gang scheduling provides shorter response time and enables interactive parallel programming. To utilize processor resources on interactive parallel programs, global state of distributed parallel...
PM : A High-Performance Communication Library for Multi-user Parallel Environments (1996)
Hiroshi Tezuka, Atsushi Hori, Yutaka Ishikawa
We have developed a communication library called PM for a workstation cluster using Sun SPARCstation 20/71's on a Myricom Myrinet. PM supports i) network context switching for the multi-user...
Yutaka Ishikawa, Mitsuhisa Sato, Junichi Shimada
Introduction High-speed networks such as Fibre Channel[1], Memory Channel[4], SCI[6], ServerNet, and Myrinet[2] are starting to be used in building parallel systems. These networks achieve a...
Design and Implementation of Metalevel Architecture in C++ - MPC++ Approach (1996)
Yutaka Ishikawa, Atsushi Hori, Mitsuhisa Sato, Motohiko Matsuda, Jörg Nolte, Hiroshi Tezuka, ...
We developed a compile-time metalevel architecture in C++, called the MPC ++ metalevel architecture, to not only extend and modify language semantics, but also extend syntax. This architecture...
This document describes a C++ template library for multi-threaded programming in MPC ++ , called MPC ++ multiple threads template library. It contains i) invoke and ainvoke function templates for...
Time Space Sharing Scheduling and Architectural Support (1995)
Atsushi Hori, Takashi Yokota, Yutaka Ishikawa, Shuichi Sakai, Hiroki Konaka, Munenori Maeda, ...
ae hori; yokota; ishikawa; sakai; konaka; m--maeda tomokiyo; jon; matsuoka; okamoto; hirono oe
RWC Massively Parallel System Software Environment (1995)
Yutaka Ishikawa, Atsushi Hori, Hiroki Konaka, Munenori Maeda, Takashi Tomokiyo, Jörg Nolte
this report, we summarize our research for the 1994 fiscal year. SCore (OS Kernel) RWC-1 (Massively Parallel Machine) MPC++
RWC Massively Parallel Software Environment and An Overview of MPC++ (1995)
Yutaka Ishikawa, Atsushi Hori, Mitsuhisa Sato, Motohiko Matsuda, Jörg Nolte, Hiroshi Tezuka, ...
this paper, we focus on MPC ++. First of all, the MPC ++ approach to a parallel programming system is described in section 2. Sections 3 and 4 show an overview of the control parallel and the MPC ++...
The Research Objective, Yutaka Ishikawa, Atsushi Hori, Hiroki Konaka, Munenori Maeda, Takashi Tomokiyo
this report, the features of MPC ++
The MPC++ Programming Language V1.0 Specification with Commentary Document Version 0.1 (1994)
This document is the reference manual of the MPC ++ programming language version 1.0 designed and developed at Tsukuba Research Center. In MPC ++ , C++ is extended for describing an operating system...
Meta-level Architecture for Extendable C++ (1994)
This document describes a meta-level architecture for the MPC ++ programming language. The MPC ++ meta-level architecture is designed so that programmers can introduce new notation and redefine...
Metalevel Decomposition in AL-1/D (1993)
Hideaki Okamura, Yutaka Ishikawa, Mario Tokoro
Research has shown that metalevel architectures and the concept of reflection are useful for modifying programming systems dynamically in a controlled way. To modify the system flexibly, it is...
An Object-Oriented Real-Time Programming Language (1992)
Yutaka Ishikawa, Mit Electrotechnical Laboratory, Hideyuki Tokuda, Clifford W. Mercer
The real-time object model is a methodology for describing real-time systems. RTC++ is a programming language that extends C++ based on this model. 66 he demand for real-time systems increases with...
Yazaki, Yoshiaki, Asukagawa, Norio, Ishikawa, Yutaka, Ohta, Eiji, Sakata, Makoto
The cytoplasmic [MgATP]/[ATP]free ratios, free Mg2+ concentrations, and phosphorylation potentials in mung bean [Vigna mungo (L.) Hepper] root tip cells were investigated by 31P nuclear magnetic...