Kazuhiko Komatsu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Kazuhiko Komatsu has a total of 63 co-authors. Here are the top co-authors by number of shared publications:
- Hiroaki Kobayashi
Graduate School of Information Sciences Tohoku University Sendai Japan
22 shared publications - Masayuki Sato
Tohoku University, Sendai, Japan
15 shared publications - Masahito Kumagai
Graduate School of Information Sciences, Tohoku University, Sendai City, Miyagi Prefecture, Japan
7 shared publications Hiroaki Kobayashi
Tohoku Univ, Grad Sch Informat Sci, 6-6-01 Aramaki Aza Aoba, Sendai, Miyagi 9808579, Japan
4 shared publications- Hiroyuki Takizawa
Tohoku University, Sendai, Japan
3 shared publications - Ilya Afanasyev
Lomonosov Moscow State University
3 shared publications
Latest Publications
- Open Access
- 25 January 2026
- Published by Association for Computing Machinery (ACM)
- p. 255-262
An analysis of memory access patterns in RISC-V vector workloads on heterogeneous memory architectures
- 25 January 2026
- Published by Association for Computing Machinery (ACM)
- p. 255-262
Open AccessAn analysis of memory access patterns in RISC-V vector workloads on heterogeneous memory architectures
- Open Access
- 15 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 96-103
Classification of Three-dimensional Electron Diffraction Data with a Large Language Model
- 15 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 96-103
Open AccessClassification of Three-dimensional Electron Diffraction Data with a Large Language Model
- Open Access
- 20 October 2025
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 37 (25-26)
Adaptive Parallelization Based on Frame‐Level and Tile‐Level Parallelisms for VVC Encoding
To meet the growing demand for high‐efficiency video compression standards, Versatile Video Coding (VVC) has been developed as a successor to High Efficiency Video Coding (HEVC). VVC offers approximately a 50% reduction in bitrate compared to HEVC while maintaining comparable visual quality. However, the improved performance of VVC comes at the cost of a significantly higher computational complexity, resulting in longer encoding times. Consequently, accelerating the VVC encoding process remains a critical challenge for its practical deployment. Leveraging the evolution of multi‐core processors and encoding tools designed for parallel processing, this study introduces an adaptive parallelization method that integrates frame‐level and tile‐level parallelisms. This method dynamically selects the number of concurrently processed frames and tiles by considering reference dependencies and their effects on coding efficiency. Moreover, the proposed method employs a content‐aware, cyclic adjustment of tile configurations to further reduce the encoding time. The evaluation results demonstrate that the proposed method can achieve a 10.58× speedup over the baseline single‐threaded implementation on average, while limiting the increase in BD‐BR to 3.23% and the degradation in BD‐PSNR to only −0.058 dB. These findings confirm that the method substantially decreases the encoding time without degrading coding efficiency. Furthermore, the results also demonstrate that the scalability of the proposed method is better than that of the conventional parallel method, Wavefront parallel processing.
- 20 October 2025
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 37 (25-26)
Open AccessAdaptive Parallelization Based on Frame‐Level and Tile‐Level Parallelisms for VVC Encoding
To meet the growing demand for high‐efficiency video compression standards, Versatile Video Coding (VVC) has been developed as a successor to High Efficiency Video Coding (HEVC). VVC offers approximately a 50% reduction in bitrate compared to HEVC while maintaining comparable visual quality. However, the improved performance of VVC comes at the cost of a significantly higher computational complexity, resulting in longer encoding times. Consequently, accelerating the VVC encoding process remains a critical challenge for its practical deployment. Leveraging the evolution of multi‐core processors and encoding tools designed for parallel processing, this study introduces an adaptive parallelization method that integrates frame‐level and tile‐level parallelisms. This method dynamically selects the number of concurrently processed frames and tiles by considering reference dependencies and their effects on coding efficiency. Moreover, the proposed method employs a content‐aware, cyclic adjustment of tile configurations to further reduce the encoding time. The evaluation results demonstrate that the proposed method can achieve a 10.58× speedup over the baseline single‐threaded implementation on average, while limiting the increase in BD‐BR to 3.23% and the degradation in BD‐PSNR to only −0.058 dB. These findings confirm that the method substantially decreases the encoding time without degrading coding efficiency. Furthermore, the results also demonstrate that the scalability of the proposed method is better than that of the conventional parallel method, Wavefront parallel processing.
- Open Access
- 20 June 2025
- Published by IOP Publishing in Quantum Science and Technology
- Vol. 10 (3) , 035042
Single photon coherent Ising machines for constrained optimization problems
A Coherent Ising machine (CIM) is an oscillator-network-based analog computing system to circumvent the bottleneck in von Neumann digital computing architectures. The CIM consists of a network of degenerate optical parametric oscillators (DOPOs) and is designed to find a ground state or perform Boltzmann sampling for all degenerate ground states and low-energy excited states in combinatorial optimization problems. A nonlinear measurement feedback scheme, called chaotic amplitude control (CAC), has recently been proposed to correct pulse amplitude inhomogeneity and thereby faithfully map the Ising Hamiltonian to the loss landscape of the DOPO network. However, the quantum limit of the CIM-CAC performance is not fully explored yet. This work clarifies how the quantum noise squeezing and the measurement-induced state shift in repeated indirect quantum measurements improve the system performance. From the numerical simulation on the Ising model with the Zeeman terms, obtained from combinatorial clustering problems formulated as constrained optimization problems, it is revealed that the CIM-CAC operating in a single photon per pulse regime dramatically outperforms the standard CIM-CAC with a large photon number per pulse. This is because the standard CIM-CAC is often trapped in a periodic trajectory and cannot escape from there. On the other hand, the significant improvement is brought by the noise-induced amplitude jump in the single photon CIM-CAC.
- 20 June 2025
- Published by IOP Publishing in Quantum Science and Technology
- Vol. 10 (3) , 035042
Open AccessSingle photon coherent Ising machines for constrained optimization problems
A Coherent Ising machine (CIM) is an oscillator-network-based analog computing system to circumvent the bottleneck in von Neumann digital computing architectures. The CIM consists of a network of degenerate optical parametric oscillators (DOPOs) and is designed to find a ground state or perform Boltzmann sampling for all degenerate ground states and low-energy excited states in combinatorial optimization problems. A nonlinear measurement feedback scheme, called chaotic amplitude control (CAC), has recently been proposed to correct pulse amplitude inhomogeneity and thereby faithfully map the Ising Hamiltonian to the loss landscape of the DOPO network. However, the quantum limit of the CIM-CAC performance is not fully explored yet. This work clarifies how the quantum noise squeezing and the measurement-induced state shift in repeated indirect quantum measurements improve the system performance. From the numerical simulation on the Ising model with the Zeeman terms, obtained from combinatorial clustering problems formulated as constrained optimization problems, it is revealed that the CIM-CAC operating in a single photon per pulse regime dramatically outperforms the standard CIM-CAC with a large photon number per pulse. This is because the standard CIM-CAC is often trapped in a periodic trajectory and cannot escape from there. On the other hand, the significant improvement is brought by the noise-induced amplitude jump in the single photon CIM-CAC.
- 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
- 18 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 45, 544-551
Appropriate Graph-Algorithm Selection for Edge Devices Using Machine Learning
As a new information technology, edge computing has attracted much attention in recent years. Edge computing collects, stores, and processes data on the edge devices, such as smartphones and sensors. Edge devices can process data in real time by reducing communication time with central servers. Therefore, various algorithms for preprocessing should be executed on edge devices to provide services rapidly. Graph algorithms are one of such candidates, because graph data play an important role in representing a variety of information around us such as maps, SNS, and web structures, which is one of data collected and processed on edge devices. However, since execution time of graph algorithms varies greatly depending on the graph data, the preprocessing on edge devices may take a long time. Therefore, it is necessary to select an appropriate algorithm and process data for edge devices as fast as possible. This paper proposes a method to select the appropriate graph algorithm on edge devices. By using machine learning that takes features of graph data as input, the performances of graph algorithms, such as the execution times, are predicted. Then, the proposed method selects a suitable algorithm for requests of edge users. The evaluation result demonstrates that the proposed method can select the appropriate algorithm from the several algorithms depending on the characteristics of graph data.
- 18 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 45, 544-551
Appropriate Graph-Algorithm Selection for Edge Devices Using Machine Learning
As a new information technology, edge computing has attracted much attention in recent years. Edge computing collects, stores, and processes data on the edge devices, such as smartphones and sensors. Edge devices can process data in real time by reducing communication time with central servers. Therefore, various algorithms for preprocessing should be executed on edge devices to provide services rapidly. Graph algorithms are one of such candidates, because graph data play an important role in representing a variety of information around us such as maps, SNS, and web structures, which is one of data collected and processed on edge devices. However, since execution time of graph algorithms varies greatly depending on the graph data, the preprocessing on edge devices may take a long time. Therefore, it is necessary to select an appropriate algorithm and process data for edge devices as fast as possible. This paper proposes a method to select the appropriate graph algorithm on edge devices. By using machine learning that takes features of graph data as input, the performances of graph algorithms, such as the execution times, are predicted. Then, the proposed method selects a suitable algorithm for requests of edge users. The evaluation result demonstrates that the proposed method can select the appropriate algorithm from the several algorithms depending on the characteristics of graph data.
- 18 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 53, 600-607
A Constraint Partition Method for Combinatorial Optimization Problems
Ising machines have attracted much attention in recent years because of their potential in solving combinatorial optimization problems that are difficult to be solved using conventional computers. Ising machines solve the optimization problem that can be transformed into a quadratic unconstrained binary optimization (QUBO) problem. In the case of a problem with constraints, named a constraint problem, an objective function and constraint function with penalty coefficients are combined into a single Hamiltonian of a QUBO problem. When solving a constraint problem, the accuracy of the solution decreases compared to a QUBO problem with an objective function only. This is because penalty coefficients are set to excessively large values to prevent constraint violations. As a result, it becomes difficult to minimize the objective function. To solve this problem, this paper proposes a method that partitions constraint functions and reduces the value of the penalty coefficients. The performance evaluation using the traveling salesperson problem (TSP) with one-hot constraints demonstrates that the proposed method can improve the solution accuracy compared to the conventional method.
- 18 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 53, 600-607
A Constraint Partition Method for Combinatorial Optimization Problems
Ising machines have attracted much attention in recent years because of their potential in solving combinatorial optimization problems that are difficult to be solved using conventional computers. Ising machines solve the optimization problem that can be transformed into a quadratic unconstrained binary optimization (QUBO) problem. In the case of a problem with constraints, named a constraint problem, an objective function and constraint function with penalty coefficients are combined into a single Hamiltonian of a QUBO problem. When solving a constraint problem, the accuracy of the solution decreases compared to a QUBO problem with an objective function only. This is because penalty coefficients are set to excessively large values to prevent constraint violations. As a result, it becomes difficult to minimize the objective function. To solve this problem, this paper proposes a method that partitions constraint functions and reduces the value of the penalty coefficients. The performance evaluation using the traveling salesperson problem (TSP) with one-hot constraints demonstrates that the proposed method can improve the solution accuracy compared to the conventional method.
- 15 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1949-1956
Multi-scale Loss based Electron Microscopic Image Pair Matching Method
Nanodiffraction Imaging (NDI), a novel imaging technique based on the scanning transmission electron mi-croscopy (STEM), helps the understanding of the relationships between micro-structure and macro-properties. However, the analysis requires image pair matching tasks through meticulous and time-consuming observation and selection by domain experts. Therefore, this paper proposes an image pair matching method for NDI images. The proposed method adopts a two-step training approach. The first step is to perform pre-training by contrastive learning on specialized NDI images. The second step is fine-tuning by a customized model on image pair matching tasks. In the second step, the training is performed with the incorporation of a special loss function, OriDist loss. This loss function is designed to focus on orientation and distribution of multi-scale features. The evaluation results demonstrate the ability of the proposed method to achieve high-accuracy NDI image pair matching, and efficiently reduce the search space of candidate matching images, resulting in a significant reduction in the human workload. Through an extensive ablation study, each component of the proposed method shows positive contributions to the overall performance.
- 15 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1949-1956
Multi-scale Loss based Electron Microscopic Image Pair Matching Method
Nanodiffraction Imaging (NDI), a novel imaging technique based on the scanning transmission electron mi-croscopy (STEM), helps the understanding of the relationships between micro-structure and macro-properties. However, the analysis requires image pair matching tasks through meticulous and time-consuming observation and selection by domain experts. Therefore, this paper proposes an image pair matching method for NDI images. The proposed method adopts a two-step training approach. The first step is to perform pre-training by contrastive learning on specialized NDI images. The second step is fine-tuning by a customized model on image pair matching tasks. In the second step, the training is performed with the incorporation of a special loss function, OriDist loss. This loss function is designed to focus on orientation and distribution of multi-scale features. The evaluation results demonstrate the ability of the proposed method to achieve high-accuracy NDI image pair matching, and efficiently reduce the search space of candidate matching images, resulting in a significant reduction in the human workload. Through an extensive ablation study, each component of the proposed method shows positive contributions to the overall performance.
- Open Access
- 9 May 2023
- Published by Association for Computing Machinery (ACM)
- p. 185-188
Performance Evaluation of Tsunami Evacuation Route Planning on Multiple Annealing Machines
This paper focuses on the performance evaluation of annealing machines based on quantum annealing and simulated annealing, i.e., D-wave advantage, Fixstars Amplify Annealing Engine, D-wave Neal, and Vector Annealing, by solving a combinatorial optimization problem, tsunami evacuation route planning. First, the problem is modeled into mathematical formulas and converted into the Quadratic Unconstrained Binary Optimization (QUBO) form as an input, and will be executed on the four annealing machines. Then, the relationship between constraint weight in the objective function and the constraint compliance conditions is also discussed. Finally, the evaluation and discussion are conducted from three aspects, the maximum number of evacuees, Time to Solution (TTS), and the solutions quality.
- 9 May 2023
- Published by Association for Computing Machinery (ACM)
- p. 185-188
Open AccessPerformance Evaluation of Tsunami Evacuation Route Planning on Multiple Annealing Machines
This paper focuses on the performance evaluation of annealing machines based on quantum annealing and simulated annealing, i.e., D-wave advantage, Fixstars Amplify Annealing Engine, D-wave Neal, and Vector Annealing, by solving a combinatorial optimization problem, tsunami evacuation route planning. First, the problem is modeled into mathematical formulas and converted into the Quadratic Unconstrained Binary Optimization (QUBO) form as an input, and will be executed on the four annealing machines. Then, the relationship between constraint weight in the objective function and the constraint compliance conditions is also discussed. Finally, the evaluation and discussion are conducted from three aspects, the maximum number of evacuees, Time to Solution (TTS), and the solutions quality.