JOPSS - Search Results

Journal Articles

High-speed classification of coherent X-ray diffraction Patterns on the K computer for high-resolution single biomolecule imaging

Tokuhisa, Atsushi*; Arai, Junya*; Jochi, Yasumasa*; Ono, Yoshiyuki*; Kameyama, Toyohisa*; Yamamoto, Keiji*; Hatanaka, Masayuki*; Gerofi, B.*; Shimada, Akio*; Kurokawa, Motoyoshi*; et al.

Journal of Synchrotron Radiation, 20(6), p.899 - 904, 2013/11

https://doi.org/10.1107/S0909049513022152

Times Cited Count：5 Percentile：28.14(Instruments & Instrumentation)

Journal Articles

MPI performance measurement on nodes of the earth simulator

Uehara, Hitoshi; Tamura, Masanori; Yokokawa, Mitsuo

Hai Pafomansu Komputingu To Keisan Kagaku Shimpojium (HPCS2002) Rombunshu, p.73 - 80, 2002/01

The Earth Simulator is an ultra high-speed supercomputer which is developed now for global environment change simulations. For achieving high performance computing on large scale distributed memory parallel computer such as the Earth Simulator, an optimization of communication processings in user applications is required, and the optimization needs an evaluation for performances of communication methods. In the Earth Simulator, Message Passing Interface (MPI) is supported as the communication method. We evaluated performances of the MPI-1/MPI-2 functions on the Earth Simulator in detail using MBL which we developed.

Journal Articles

A Simulation on homogeneous isotropic turbulent flows on the earth simulator

Yokokawa, Mitsuo; Saito, Minoru*; Ishihara, Takashi*; Kaneda, Yukio*

Hai Pafomansu Komputingu To Keisan Kagaku Shimpojium (HPCS2002) Rombunshu, p.125 - 131, 2002/01

With the advent of supercomputers, large-scale direct numerical simulations(DNS) of the Navier-Stokes equations are possible to carry out. However, larger scale DNS is required to make turbulence clear and to make a parameterization of turbulence. The Earth Simulator is a distributed memory parallel supercomputer whose peak speed is 40 Tflop/s and it is very useful to make a very large-scale DNS. Simulation codes named have been developed for the Earth Simulator and their correctness have been validated by compareing the simulation results by with ones by a conventinal code. The sustatined performance of 3.72Gflop/s is obtained in the execution of the serial version with the problem size of on an arithmetic processor(AP) of the Earth Simulator. Speedup of about 7 is achieved in 8APs by the parallelized code. Sppedup is degraded by the execution of 8 processor nodes, however, efficiency of 25% against the peak performance is obtained for the problem size of .

Journal Articles

Vector performance prediction of kernel loops on Earth Simulator

Yokokawa, Mitsuo; Saito, Minoru*; Hagiwara, Takashi*; Isobe, Yoko*; Jinguji, Satoshi*

Nihon Keisan Kogakkai Rombunshu, 4, p.31 - 36, 2002/00

http://joi.jlc.jst.go.jp/JST.JSTAGE/jsces/2001.20010040

Earth simulator is a distributed memory parallel system which consists of 640 processor nodes connected by a full crossbar network. Each processor node is a shared memory system which is composed of eight vector processors. The total peak performance and main memory capacity are 40Tflops and 10TB, respectively. A performance prediction system GS for the Earth Simulator has been developed to estimate sustained performance of programs. To validate accuracy of vector performance prediction by the GS, the processing times for three groups of kernel loops estimated by the GS are compared with the ones measured on SX-4. It is found that the absolute relative errors of the processing time are 0.89%,1.42% and 6.81% in average for three groups. The sustained performance of three groups on a processor of the Earth Simulator have been estimated by the GS and those performance are 5.94Gflops,3.76Gflops and 2.17Gflops in average.

Journal Articles

High speed calculation for solid molecular dynamics on vector processors

Itakura, Kenichi; Yokokawa, Mitsuo; Shimizu, Futoshi; Kimizuka, Hajime*; Kaburaki, Hideo

Joho Shori Gakkai Kenkyu Hokoku 2001-HPC-88, p.67 - 72, 2001/10

he Earth Simulator which is under development has 640 processor nodes and its peak performance is 40 Tflop/s. In this study, we have evaluated performance of solid molecular dynamics simulation on an SMP node of the Earth Simulator. In molecular dynamics simulation, each particle is influenced by all particles within a cut-off region and the representation of these pairs of particles is made by a matrix. Two matrix representations, compressed row form and jagged diagonal form, are considered for vectorization. The jagged diagonal form is better than the compressed row form in performance on a vector processor for the force calculation of every pairs, because the vector length of the former is longer than that of the latter. However, computational cost for converting the normal matrix form to the jagged diagonal form is quite expensive and the total performance in using the jagged diagonal form is low. Speedup by parallelization with the compressed row form is 2.4 to 2.7 with 8 vector processors.

Journal Articles

MBL2: MPI benchmark program library for MPI-2

Uehara, Hitoshi; Tsuda, Yoshinori*; Yokokawa, Mitsuo

Joho Shori Gakkai Kenkyu Hokoku 2001-HPC-87, 2001(77), p.67 - 72, 2001/07

MPI is one of major message communication interfaces for application programs. The MPI consists of an MPI-1 as a basic specification, and an MPI-2 as extensions.Some benchmark program for MPI-1 had been proposed already. However benchmark programs for MPI-2 are a little and their measurements are limited. We have developed an MPI benchmark program library for MPI-2 (MBL2) which measures the detail performance of MPI-I/O and RMA functions of MPI-2. In this report, we describe the MBL2 and performance data of MPI-2 on VPP5000, which we measured using MBL2.

Journal Articles

MBL2; MPI benchmark program library for MPI-2

Uehara, Hitoshi; Tsuda, Yoshinori*; Yokokawa, Mitsuo

Heiretsu Shori Shimpojiumu (JSPP2001) Rombunshu, 2001(6), p.91 - 92, 2001/06

MPI in one of major message communication interfaces for application programs. The MPI consists of an MPI-1 as a basic specification, and an MPI-2 as extensions. Some benchmark programs for MPI-1 capabilities are proposed. However benchmark programs for MPI-2 are a little and their measurements are limited. We have developed an MPI benchmark program library for MPI-2 (MBL2) which measures the performance of MPI-I/O and RMA.

Journal Articles