Mitsuo Yokokawa
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Mitsuo Yokokawa has a total of 146 co-authors. Here are the top co-authors by number of shared publications:
Atsuya Uno
RIKEN Center for Computational Science, Kobe650-0047, Japan
11 shared publicationsTakashi Ishihara
8 shared publicationsYukio Kaneda
Nagoya University
7 shared publicationsHideo Kaburaki
5 shared publications- Takashi Ishihara
Okayama University
5 shared publications Tadashi Watanabe
4 shared publications
Latest Publications
- 30 June 2024
- Published by Springer Nature
- p. 173-187
File I/O Cache Performance of Supercomputer Fugaku Using an Out-of-Core Direct Numerical Simulation Code of Turbulence
- 30 June 2024
- Published by Springer Nature
- p. 173-187
File I/O Cache Performance of Supercomputer Fugaku Using an Out-of-Core Direct Numerical Simulation Code of Turbulence
- 27 June 2024
- Published by Springer Nature
- p. 383-397
An Asymptotic Parallel Linear Solver and Its Application to Direct Numerical Simulation for Compressible Turbulence
- 27 June 2024
- Published by Springer Nature
- p. 383-397
An Asymptotic Parallel Linear Solver and Its Application to Direct Numerical Simulation for Compressible Turbulence
- 29 April 2023
- Published by Elsevier in Computers & Fluids
- Vol. 261
Performance evaluation of parallel direct numerical simulation code on supercomputer SX-Aurora TSUBASA
- 29 April 2023
- Published by Elsevier in Computers & Fluids
- Vol. 261
Performance evaluation of parallel direct numerical simulation code on supercomputer SX-Aurora TSUBASA
- 25 October 2021
- Published by Cambridge University Press (CUP) in Journal of Fluid Mechanics
- Vol. 929, A1
Statistics of local Reynolds number in box turbulence: ratio of inertial to viscous forces
In high-Reynolds-number turbulence the spatial distribution of velocity fluctuation at small scales is strongly non-uniform. In accordance with the non-uniformity, the distributions of the inertial and viscous forces are also non-uniform. According to direct numerical simulation (DNS) of forced turbulence of an incompressible fluid obeying the Navier–Stokes equation in a periodic box at the Taylor microscale Reynolds number $R_\lambda \approx 1100$ , the average $\langle R_{loc}\rangle$ over the space of the ‘local Reynolds number’ $R_ {loc}$ , which is defined as the ratio of inertial to viscous forces at each point in the flow, is much smaller than the conventional ‘Reynolds number’ given by $Re \equiv UL/\nu$ , where $U$ and $L$ are the characteristic velocity and length of the energy-containing eddies, and $\nu$ is the kinematic viscosity. While both conditional averages of the inertial and viscous forces for a given squared vorticity $\omega ^{2}$ increase with $\omega ^{2}$ at large $\omega ^{2}$ , the conditional average of $R_ {loc}$ is almost independent of $\omega ^{2}$ . A comparison of the DNS field with a random structureless velocity field suggests that the increase in the conditional average of $R_ {loc}$ with $\omega ^{2}$ at large $\omega ^{2}$ is suppressed by the Navier–Stokes dynamics. Something similar is also true for the conditional averages for a given local energy dissipation rate per unit mass. Certain features of intermittency effects such as that on the $Re$ dependence of $\langle R_{loc}\rangle$ are explained by a multi-fractal model by Dubrulle (J. Fluid Mech., vol. 867, 2019, P1).- 25 October 2021
- Published by Cambridge University Press (CUP) in Journal of Fluid Mechanics
- Vol. 929, A1
Statistics of local Reynolds number in box turbulence: ratio of inertial to viscous forces
In high-Reynolds-number turbulence the spatial distribution of velocity fluctuation at small scales is strongly non-uniform. In accordance with the non-uniformity, the distributions of the inertial and viscous forces are also non-uniform. According to direct numerical simulation (DNS) of forced turbulence of an incompressible fluid obeying the Navier–Stokes equation in a periodic box at the Taylor microscale Reynolds number $R_\lambda \approx 1100$ , the average $\langle R_{loc}\rangle$ over the space of the ‘local Reynolds number’ $R_ {loc}$ , which is defined as the ratio of inertial to viscous forces at each point in the flow, is much smaller than the conventional ‘Reynolds number’ given by $Re \equiv UL/\nu$ , where $U$ and $L$ are the characteristic velocity and length of the energy-containing eddies, and $\nu$ is the kinematic viscosity. While both conditional averages of the inertial and viscous forces for a given squared vorticity $\omega ^{2}$ increase with $\omega ^{2}$ at large $\omega ^{2}$ , the conditional average of $R_ {loc}$ is almost independent of $\omega ^{2}$ . A comparison of the DNS field with a random structureless velocity field suggests that the increase in the conditional average of $R_ {loc}$ with $\omega ^{2}$ at large $\omega ^{2}$ is suppressed by the Navier–Stokes dynamics. Something similar is also true for the conditional averages for a given local energy dissipation rate per unit mass. Certain features of intermittency effects such as that on the $Re$ dependence of $\langle R_{loc}\rangle$ are explained by a multi-fractal model by Dubrulle (J. Fluid Mech., vol. 867, 2019, P1).- 8 June 2019
- Published by Springer Nature
- p. 587-595
Performance of a Two-Path Aliasing Free Calculation of a Spectral DNS Code
- 8 June 2019
- Published by Springer Nature
- p. 587-595
Performance of a Two-Path Aliasing Free Calculation of a Spectral DNS Code
- 1 November 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 685-696
Performance Evaluation of a Vector Supercomputer SX-Aurora TSUBASA
A new SX-Aurora TSUBASA vector supercomputer has been released, and it features a new system architecture and a new execution model to achieve high sustained performance, especially for memory-intensive applications. In SX-Aurora TSUBASA, the vector host (VH) of a standard ×86 Linux node is attached to the vector engine (VE) of the newly developed vector processor. An application is executed on the VE, and only system calls are offloaded to the VH. This new execution model can avoid redundant data transfers between the VH and VE that can easily become a bottleneck in the conventional execution model. This paper examines the potential of SX-Aurora TSUBASA. First, the basic performance is clarified by evaluating benchmark programs. Then, the effectiveness of the new execution model is examined by using a microbenchmark. Finally, the potential of SX-Aurora TSUBASA is clarified through evaluations of practical applications.
- 1 November 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 685-696
Performance Evaluation of a Vector Supercomputer SX-Aurora TSUBASA
A new SX-Aurora TSUBASA vector supercomputer has been released, and it features a new system architecture and a new execution model to achieve high sustained performance, especially for memory-intensive applications. In SX-Aurora TSUBASA, the vector host (VH) of a standard ×86 Linux node is attached to the vector engine (VE) of the newly developed vector processor. An application is executed on the VE, and only system calls are offloaded to the VH. This new execution model can avoid redundant data transfers between the VH and VE that can easily become a bottleneck in the conventional execution model. This paper examines the potential of SX-Aurora TSUBASA. First, the basic performance is clarified by evaluating benchmark programs. Then, the effectiveness of the new execution model is examined by using a microbenchmark. Finally, the potential of SX-Aurora TSUBASA is clarified through evaluations of practical applications.
- 1 November 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 257-262
Preconditioner Auto-Tuning Using Deep Learning for Sparse Iterative Algorithms
In numerical libraries for sparse matrix operations, there are many tuning parameters related to implementation selection. Selection of different tuning parameters could result in totally different performance. Moreover, optimal implementation depends on the sparse matrices to be operated. It is difficult to find optimal implementation without executing each implementation and thereby examining its performance on a given sparse matrix. In this study, we propose an implementation selection method for sparse iterative algorithms and preconditioners in a numerical library using deep learning. The proposed method uses full color images to represent the features of a sparse matrix. We present an image generation method for partitioning a given matrix (to generate its feature image) so that the value of each matrix element is considered in the implementation selection. We then evaluate the effectiveness of the proposed method by conducting a numerical experiment. In this experiment, the accuracy of implementation selection is evaluated. The training data comprise a pair of sparse matrix and its optimal implementation. The optimal implementation of each sparse matrix in the training data is obtained in advance by executing every implementation and getting the best one. The experimental results obtained using the proposed method show that the accuracy of selecting the optimal implementation of each sparse matrix is 79.5%.
- 1 November 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 257-262
Preconditioner Auto-Tuning Using Deep Learning for Sparse Iterative Algorithms
In numerical libraries for sparse matrix operations, there are many tuning parameters related to implementation selection. Selection of different tuning parameters could result in totally different performance. Moreover, optimal implementation depends on the sparse matrices to be operated. It is difficult to find optimal implementation without executing each implementation and thereby examining its performance on a given sparse matrix. In this study, we propose an implementation selection method for sparse iterative algorithms and preconditioners in a numerical library using deep learning. The proposed method uses full color images to represent the features of a sparse matrix. We present an image generation method for partitioning a given matrix (to generate its feature image) so that the value of each matrix element is considered in the implementation selection. We then evaluate the effectiveness of the proposed method by conducting a numerical experiment. In this experiment, the accuracy of implementation selection is evaluated. The training data comprise a pair of sparse matrix and its optimal implementation. The optimal implementation of each sparse matrix in the training data is obtained in advance by executing every implementation and getting the best one. The experimental results obtained using the proposed method show that the accuracy of selecting the optimal implementation of each sparse matrix is 79.5%.