Yutaka Ishikawa
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
Research Interests
Yutaka Ishikawa has a total of 56 co-authors. Here are the top co-authors by number of shared publications:
- Balazs Gerofi
Intel Corporation, Hillsboro, Oregon, USA
7 shared publications - Robert Wiśniewski
Intel Corporation
4 shared publications - Atsushi Hori
National Institute of Informatics
2 shared publications - Kaiming Ouyang
NVIDIA Corporation
2 shared publications Kengo Nakajima
The University of Tokyo, Japan
2 shared publications- Min Si
Meta, Menlo Park, USA
2 shared publications
Latest Publications
- Open Access
- 1 September 2025
- Published by Association for Computing Machinery (ACM) in ACM Computing Surveys
- Vol. 58 (1) , 1-34
A Survey on the Shared Address Space with Privatized Static Variables (SAS-PSV) Execution Model for the Many-Core Era
Parallel applications often use MPI processes and OpenMP threads. Those parallel execution models, multi-process and multi-thread, were invented to increase efficiency on uniprocessor systems. In the multi-process approach, each process’s isolated address space may make communication expensive; in the multi-thread design, shared variables may cause access conflicts and stall executions. Processes or threads interact and exchange information more often as CPU cores increase, and traditional execution models may create bottlenecks. The paradigm shift from uniprocessor to many-core systems necessitates the development of new parallel execution models to address challenges posed by the two parallel models. When processes share an address space, what happens? If threads don’t share static variables? Sharing an address space and privatizing static variables reduces information exchange and shared static variable exclusion costs. This survey investigates SAS-PSV (SAS-PSV) , a new execution architecture that allows shared address space and static variable privatization. This notion is implemented by MPC, SMARTMAP, PVAS, PiP, and AMPI. Each has a different approach and execution. This article analyzes these implementations’ concepts, details, and hidden defects. We also present SAS-PSV applications and issues that need to be solved.
- 1 September 2025
- Published by Association for Computing Machinery (ACM) in ACM Computing Surveys
- Vol. 58 (1) , 1-34
Open AccessA Survey on the Shared Address Space with Privatized Static Variables (SAS-PSV) Execution Model for the Many-Core Era
Parallel applications often use MPI processes and OpenMP threads. Those parallel execution models, multi-process and multi-thread, were invented to increase efficiency on uniprocessor systems. In the multi-process approach, each process’s isolated address space may make communication expensive; in the multi-thread design, shared variables may cause access conflicts and stall executions. Processes or threads interact and exchange information more often as CPU cores increase, and traditional execution models may create bottlenecks. The paradigm shift from uniprocessor to many-core systems necessitates the development of new parallel execution models to address challenges posed by the two parallel models. When processes share an address space, what happens? If threads don’t share static variables? Sharing an address space and privatizing static variables reduces information exchange and shared static variable exclusion costs. This survey investigates SAS-PSV (SAS-PSV) , a new execution architecture that allows shared address space and static variable privatization. This notion is implemented by MPC, SMARTMAP, PVAS, PiP, and AMPI. Each has a different approach and execution. This article analyzes these implementations’ concepts, details, and hidden defects. We also present SAS-PSV applications and issues that need to be solved.
- 16 November 2024
- Published by Springer Nature
- p. 289-297
Distributed Dataset Framework for Large Language Models Pre-training
- 16 November 2024
- Published by Springer Nature
- p. 289-297
Distributed Dataset Framework for Large Language Models Pre-training
- Open Access
- 11 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-10
Big Data Assimilation: Real-time 30-second-refresh Heavy Rain Forecast Using Fugaku During Tokyo Olympics and Paralympics
Real-time 30-second-refresh numerical weather prediction (NWP) was performed with exclusive use of 11,580 nodes (~7%) of supercomputer Fugaku during Tokyo Olympics and Paralympics in 2021. Total 75,248 forecasts were disseminated in the 1-month period mostly stably with time-to-solution less than 3 minutes for 30-minute forecast. Japan's Big Data Assimilation (BDA) project developed the novel NWP system for precise prediction of hazardous rains toward solving the global climate crisis. Compared with typical 1-hour-refresh systems, the BDA system offered two orders of magnitude increase in problem size and revealed the effectiveness of 30-second refresh for highly nonlinear, rapidly evolving convective rains. To achieve the required time-to-solution for real-time 30-second refresh with high accuracy, the core BDA software incorporated single precision and enhanced parallel I/O with properly selected configurations of 1000 ensemble members and 500-m-mesh weather model. The massively parallel, I/O intensive real-time BDA computation demonstrated a promising future direction.
- 11 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-10
Open AccessBig Data Assimilation: Real-time 30-second-refresh Heavy Rain Forecast Using Fugaku During Tokyo Olympics and Paralympics
Real-time 30-second-refresh numerical weather prediction (NWP) was performed with exclusive use of 11,580 nodes (~7%) of supercomputer Fugaku during Tokyo Olympics and Paralympics in 2021. Total 75,248 forecasts were disseminated in the 1-month period mostly stably with time-to-solution less than 3 minutes for 30-minute forecast. Japan's Big Data Assimilation (BDA) project developed the novel NWP system for precise prediction of hazardous rains toward solving the global climate crisis. Compared with typical 1-hour-refresh systems, the BDA system offered two orders of magnitude increase in problem size and revealed the effectiveness of 30-second refresh for highly nonlinear, rapidly evolving convective rains. To achieve the required time-to-solution for real-time 30-second refresh with high accuracy, the core BDA software incorporated single precision and enhanced parallel I/O with properly selected configurations of 1000 ensemble members and 500-m-mesh weather model. The massively parallel, I/O intensive real-time BDA computation demonstrated a promising future direction.
- Open Access
- 1 July 2022
- Published by Springer Nature
- p. 59-78
On the Difference Between Shared Memory and Shared Address Space in HPC Communication
Shared memory mechanisms, e.g., POSIX shmem or XPMEM, are widely used to implement efficient intra-node communication among processes running on the same node. While POSIX shmem allows other processes to access only newly allocated memory, XPMEM allows accessing any existing data and thus enables more efficient communication because the send buffer content can directly be copied to the receive buffer. Recently, the shared address space model has been proposed, where processes on the same node are mapped into the same address space at the time of process creation, allowing processes to access any data in the shared address space. Process-in-Process (PiP) is an implementation of such mechanism. The functionalities of shared memory mechanisms and the shared address space model look very similar – both allow accessing the data of other processes –, however, the shared address space model includes the shared memory model. Their internal mechanisms are also notably different. This paper clarifies the differences between the shared memory and the shared address space models, both qualitatively and quantitatively. This paper is not to showcase applications of the shared address space model, but through minimal modifications to an existing MPI implementation it highlights the basic differences between the two models. The following four MPI configurations are evaluated and compared; 1) POSIX Shmem, 2) XPMEM, 3) PiP-Shmem, where intra-node communication is implemented to utilize POSIX shmem but MPI processes share the same address space, and 4) PiP-XPMEM, where XPMEM functions are implemented by the PiP library (without the need for linking to XPMEM library). Evaluation is done using the Intel MPI benchmark suite and six HPC benchmarks (HPCCG, miniGhost, LULESH2.0, miniMD, miniAMR and mpiGraph). Most notably, mpiGraph performance of PiP-XPMEM outperforms the XPMEM implementation by almost 1.5x. The performance numbers of HPCCG, miniGhost, miniMD, LULESH2.0 running with PiP-Shmem and PiP-XPMEM are comparable with those of POSIX Shmem and XPMEM. PiP is not only a practical implementation of the shared address space model, but it also provides opportunities for developing new optimization techniques, which the paper further elaborates on.
- 1 July 2022
- Published by Springer Nature
- p. 59-78
Open AccessOn the Difference Between Shared Memory and Shared Address Space in HPC Communication
Shared memory mechanisms, e.g., POSIX shmem or XPMEM, are widely used to implement efficient intra-node communication among processes running on the same node. While POSIX shmem allows other processes to access only newly allocated memory, XPMEM allows accessing any existing data and thus enables more efficient communication because the send buffer content can directly be copied to the receive buffer. Recently, the shared address space model has been proposed, where processes on the same node are mapped into the same address space at the time of process creation, allowing processes to access any data in the shared address space. Process-in-Process (PiP) is an implementation of such mechanism. The functionalities of shared memory mechanisms and the shared address space model look very similar – both allow accessing the data of other processes –, however, the shared address space model includes the shared memory model. Their internal mechanisms are also notably different. This paper clarifies the differences between the shared memory and the shared address space models, both qualitatively and quantitatively. This paper is not to showcase applications of the shared address space model, but through minimal modifications to an existing MPI implementation it highlights the basic differences between the two models. The following four MPI configurations are evaluated and compared; 1) POSIX Shmem, 2) XPMEM, 3) PiP-Shmem, where intra-node communication is implemented to utilize POSIX shmem but MPI processes share the same address space, and 4) PiP-XPMEM, where XPMEM functions are implemented by the PiP library (without the need for linking to XPMEM library). Evaluation is done using the Intel MPI benchmark suite and six HPC benchmarks (HPCCG, miniGhost, LULESH2.0, miniMD, miniAMR and mpiGraph). Most notably, mpiGraph performance of PiP-XPMEM outperforms the XPMEM implementation by almost 1.5x. The performance numbers of HPCCG, miniGhost, miniMD, LULESH2.0 running with PiP-Shmem and PiP-XPMEM are comparable with those of POSIX Shmem and XPMEM. PiP is not only a practical implementation of the shared address space model, but it also provides opportunities for developing new optimization techniques, which the paper further elaborates on.
- Open Access
- 29 January 2022
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Storage
- Vol. 18 (1) , 1-25
Pattern-Based Prefetching with Adaptive Cache Management Inside of Solid-State Drives
This article proposes a pattern-based prefetching scheme with the support of adaptive cache management, at the flash translation layer of solid-state drives ( SSDs ). It works inside of SSDs and has features of OS dependence and uses transparency. Specifically, it first mines frequent block access patterns that reflect the correlation among the occurred I/O requests. Then, it compares the requests in the current time window with the identified patterns to direct prefetching data into the cache of SSDs. More importantly, to maximize the cache use efficiency, we build a mathematical model to adaptively determine the cache partition on the basis of I/O workload characteristics, for separately buffering the prefetched data and the written data. Experimental results show that our proposal can yield improvements on average read latency by 1.8 %– 36.5 % without noticeably increasing the write latency, in contrast to conventional SSD-inside prefetching schemes.
- 29 January 2022
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Storage
- Vol. 18 (1) , 1-25
Open AccessPattern-Based Prefetching with Adaptive Cache Management Inside of Solid-State Drives
This article proposes a pattern-based prefetching scheme with the support of adaptive cache management, at the flash translation layer of solid-state drives ( SSDs ). It works inside of SSDs and has features of OS dependence and uses transparency. Specifically, it first mines frequent block access patterns that reflect the correlation among the occurred I/O requests. Then, it compares the requests in the current time window with the identified patterns to direct prefetching data into the cache of SSDs. More importantly, to maximize the cache use efficiency, we build a mathematical model to adaptively determine the cache partition on the basis of I/O workload characteristics, for separately buffering the prefetched data and the written data. Experimental results show that our proposal can yield improvements on average read latency by 1.8 %– 36.5 % without noticeably increasing the write latency, in contrast to conventional SSD-inside prefetching schemes.
- Open Access
- 20 January 2021
- Published by Association for Computing Machinery (ACM)
- p. 46-49
Efficient Parallel Multigrid Method on Intel Xeon Phi Clusters
The parallel multigrid method is expected to play an important role in scientific computing on exa-scale supercomputer systems for solving large-scale linear equations with sparse matrices. Because solving sparse linear systems is a very memory-bound process, efficient method for storage of coefficient matrices is a crucial issue. In the previous works, authors implemented sliced ELL method to parallel conjugate gradient solvers with multigrid preconditioning (MGCG) for the application on 3D groundwater flow through heterogeneous porous media (pGW3D-FVM), and excellent performance has been obtained on large-scale multicore/manycore clusters. In the present work, authors introduced SELL-C-σ to the MGCG solver, and evaluated the performance of the solver with various types of OpenMP/MPI hybrid parallel programing models on the Oakforest-PACS (OFP) system at JCAHPC using up to 1,024 nodes of Intel Xeon Phi. Because SELL-C-σ is suitable for wide-SIMD architecture, such as Xeon Phi, improvement of the performance over the sliced ELL was more than 20%. This is one of the first examples of SELL-C-σ applied to forward/backward substitutions in ILU-type smoother of multigrid solver. Furthermore, effects of IHK/McKernel has been investigated, and it achieved 11% improvement on 1,024 nodes.
- 20 January 2021
- Published by Association for Computing Machinery (ACM)
- p. 46-49
Open AccessEfficient Parallel Multigrid Method on Intel Xeon Phi Clusters
The parallel multigrid method is expected to play an important role in scientific computing on exa-scale supercomputer systems for solving large-scale linear equations with sparse matrices. Because solving sparse linear systems is a very memory-bound process, efficient method for storage of coefficient matrices is a crucial issue. In the previous works, authors implemented sliced ELL method to parallel conjugate gradient solvers with multigrid preconditioning (MGCG) for the application on 3D groundwater flow through heterogeneous porous media (pGW3D-FVM), and excellent performance has been obtained on large-scale multicore/manycore clusters. In the present work, authors introduced SELL-C-σ to the MGCG solver, and evaluated the performance of the solver with various types of OpenMP/MPI hybrid parallel programing models on the Oakforest-PACS (OFP) system at JCAHPC using up to 1,024 nodes of Intel Xeon Phi. Because SELL-C-σ is suitable for wide-SIMD architecture, such as Xeon Phi, improvement of the performance over the sliced ELL was more than 20%. This is one of the first examples of SELL-C-σ applied to forward/backward substitutions in ILU-type smoother of multigrid solver. Furthermore, effects of IHK/McKernel has been investigated, and it achieved 11% improvement on 1,024 nodes.
- 16 October 2019
- Published by Springer Nature
- p. 223-226
A New Age: An Overview of Multi-kernels
- 16 October 2019
- Published by Springer Nature
- p. 3-15
Introduction to HPC Operating Systems
- 16 October 2019
- Published by Springer Nature
- p. 19-21
Overview: The Birth of Lightweight Kernels
- 1 May 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 116-125
Performance and Scalability of Lightweight Multi-kernel Based Operating Systems
Multi-kernels leverage today's multi-core chips to run multiple operating system (OS) kernels, typically a Light Weight Kernel (LWK) and a Linux kernel, simultaneously. The LWK provides high performance and scalability, while the Linux kernel provides compatibility. Multi-kernels show the promise of being able to meet tomorrow's extreme-scale computing needs while providing strong isolation, yielding high performance and scalability needed by classical HPC applications. McKernel and mOS started as independent research initiatives to explore the above potential. Previous work described their design and architecture advantages. This paper deploys the two LWKs and presents results from running them on a 2,048-node system with Intel Xeon Phi processors (KNL) connected by Intel Omni-Path Fabric. We compare the performance of McKernel, mOS, and Linux. Although the two multi-kernel efforts approached the problem from different angles, the results show a median performance improvement of 9% with some applications as high as 280% validating the efficacy of the multi-kernel approach. We provide insight into the performance improvements and discuss the strengths of the two different multi-kernel approaches.
- 1 May 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 116-125
Performance and Scalability of Lightweight Multi-kernel Based Operating Systems
Multi-kernels leverage today's multi-core chips to run multiple operating system (OS) kernels, typically a Light Weight Kernel (LWK) and a Linux kernel, simultaneously. The LWK provides high performance and scalability, while the Linux kernel provides compatibility. Multi-kernels show the promise of being able to meet tomorrow's extreme-scale computing needs while providing strong isolation, yielding high performance and scalability needed by classical HPC applications. McKernel and mOS started as independent research initiatives to explore the above potential. Previous work described their design and architecture advantages. This paper deploys the two LWKs and presents results from running them on a 2,048-node system with Intel Xeon Phi processors (KNL) connected by Intel Omni-Path Fabric. We compare the performance of McKernel, mOS, and Linux. Although the two multi-kernel efforts approached the problem from different angles, the results show a median performance improvement of 9% with some applications as high as 280% validating the efficacy of the multi-kernel approach. We provide insight into the performance improvements and discuss the strengths of the two different multi-kernel approaches.