Hayato Yamaki
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
Hayato Yamaki has a total of 24 co-authors. Here are the top co-authors by number of shared publications:
- Hiroki Honda
The University of Electro-Communications, Chofu, Tokyo, Japan
6 shared publications - Shinobu Miwa
The University of Electro-Communications, Chofu, Tokyo, Japan
6 shared publications - Kento Sato
RIKEN Center for Computational Science, Kobe, Japan
2 shared publications - Kohei Yoshida
Department of Computer and Network Engineering, The University of Electro-Communications, Tokyo, Japan
2 shared publications - Tongxin Yang
Univ. of Electro-Communications, Chofu, Japan
2 shared publications - Abhinav Bhatele
Department of Computer Science, University of Maryland, College Park, MD, USA
1 shared publication
Latest Publications
- Open Access
- 25 January 2026
- Published by Association for Computing Machinery (ACM)
- p. 22-30
Combining System- and User-Level Approaches to Improving Energy Efficiency in GPU-Based Supercomputers
- 25 January 2026
- Published by Association for Computing Machinery (ACM)
- p. 22-30
Open AccessCombining System- and User-Level Approaches to Improving Energy Efficiency in GPU-Based Supercomputers
- 24 November 2025
- Published by Springer Nature
- p. 639-652
Analysis of MPI Parallel Code Generated by GPT-4o
- 24 November 2025
- Published by Springer Nature
- p. 639-652
Analysis of MPI Parallel Code Generated by GPT-4o
- Open Access
- 12 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 36 (8) , 1713-1727
VAHRM: Variation-Aware Resource Management in Heterogeneous Supercomputing Systems
In this paper, we propose a novel resource management technique for heterogeneous supercomputing systems affected by manufacturing variability. Our proposed technique called VAHRM (Variation-Aware Heterogeneous Resource Management) takes a holistic approach to job scheduling on highly heterogeneous computing resources. VAHRM preferentially allocates energy-efficient computing resources to an energy-consuming job in a job queue, considering the impact on both the job turnaround time and the power consumption of individual resources. Furthermore, we have developed a novel approach to modeling the power consumption of computing resources that have manufacturing variability. Our approach called TSMVA (Two-Stage Modeling with Variation Awareness) enables us to generate the first variation-aware GPU power models, which can correctly estimate the power consumption of each GPU for a given job. Our experimental results show that, compared to conventional first-come-first-serve (FCFS) and state-of-the-art variation-aware scheduling algorithms, VAHRM can achieve respective improvements in system energy efficiency of up to 5.8% and 5.4% (4.5% and 4.2% on average) while reducing the average turnaround time of 21.2% and 11.9%, respectively, for various workloads obtained from a production system.
- 12 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 36 (8) , 1713-1727
Open AccessVAHRM: Variation-Aware Resource Management in Heterogeneous Supercomputing Systems
In this paper, we propose a novel resource management technique for heterogeneous supercomputing systems affected by manufacturing variability. Our proposed technique called VAHRM (Variation-Aware Heterogeneous Resource Management) takes a holistic approach to job scheduling on highly heterogeneous computing resources. VAHRM preferentially allocates energy-efficient computing resources to an energy-consuming job in a job queue, considering the impact on both the job turnaround time and the power consumption of individual resources. Furthermore, we have developed a novel approach to modeling the power consumption of computing resources that have manufacturing variability. Our approach called TSMVA (Two-Stage Modeling with Variation Awareness) enables us to generate the first variation-aware GPU power models, which can correctly estimate the power consumption of each GPU for a given job. Our experimental results show that, compared to conventional first-come-first-serve (FCFS) and state-of-the-art variation-aware scheduling algorithms, VAHRM can achieve respective improvements in system energy efficiency of up to 5.8% and 5.4% (4.5% and 4.2% on average) while reducing the average turnaround time of 21.2% and 11.9%, respectively, for various workloads obtained from a production system.
- Open Access
- 20 January 2025
- Published by Association for Computing Machinery (ACM)
- p. 1350-1356
CACTI-CNFET: an Analytical Tool for Timing, Power, and Area of SRAMs with Carbon Nanotube Field Effect Transistors
Carbon nanotube field effect transistors (CNFETs) are expected to replace silicon-based metal oxide semiconductor field effect transistors (MOSFETs) to improve the power efficiency and performance of microprocessors. However, the design of CNFET processors is not as mature as that of silicon-based processors because there is no architecture-level analytical tool for CNFET processors. Since circuit-level analysis such as RTL analysis is a very time-consuming and troublesome task, architecture-level analysis is needed for the rapid design of processors optimized for CNFETs. In this paper, we propose CACTI-CNFET, the first architecture-level analytical tool for SRAMs designed with CNFETs. Similar to CACTI, a widely used tool for analyzing silicon-based SRAMs, CACTI-CNFET allows us to quickly estimate the timing, power, and area of CNFET SRAMs only with architecture-level descriptions such as memory capacity and port count, thus helping us make a quick decision on the architecture of various SRAMs used in a targeted processor. Because CACTI-CNFET is implemented on top of CACTI and published as an open source project, all CACTI users can use and extend this new functionality freely. Our experimental results show that CACTI-CNFET can estimate the leakage power, dynamic energy consumption for write and read operations, and circuit delay of various CNFET SRAMs with average errors of 7.40%, 7.08%, 2.36%, and 8.33%, respectively. Moreover, our experimental results show that CACTI-CNFET can estimate the area of a CNFET SRAM with an error of 10.25%.
- 20 January 2025
- Published by Association for Computing Machinery (ACM)
- p. 1350-1356
Open AccessCACTI-CNFET: an Analytical Tool for Timing, Power, and Area of SRAMs with Carbon Nanotube Field Effect Transistors
Carbon nanotube field effect transistors (CNFETs) are expected to replace silicon-based metal oxide semiconductor field effect transistors (MOSFETs) to improve the power efficiency and performance of microprocessors. However, the design of CNFET processors is not as mature as that of silicon-based processors because there is no architecture-level analytical tool for CNFET processors. Since circuit-level analysis such as RTL analysis is a very time-consuming and troublesome task, architecture-level analysis is needed for the rapid design of processors optimized for CNFETs. In this paper, we propose CACTI-CNFET, the first architecture-level analytical tool for SRAMs designed with CNFETs. Similar to CACTI, a widely used tool for analyzing silicon-based SRAMs, CACTI-CNFET allows us to quickly estimate the timing, power, and area of CNFET SRAMs only with architecture-level descriptions such as memory capacity and port count, thus helping us make a quick decision on the architecture of various SRAMs used in a targeted processor. Because CACTI-CNFET is implemented on top of CACTI and published as an open source project, all CACTI users can use and extend this new functionality freely. Our experimental results show that CACTI-CNFET can estimate the leakage power, dynamic energy consumption for write and read operations, and circuit delay of various CNFET SRAMs with average errors of 7.40%, 7.08%, 2.36%, and 8.33%, respectively. Moreover, our experimental results show that CACTI-CNFET can estimate the area of a CNFET SRAM with an error of 10.25%.
- Open Access
- 7 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 165335-165347
CNFET-OCL: Open-Source Cell Libraries for Advanced CNFET Technologies
In this paper, we propose CNFET-OCL, the first open-source cell libraries for 5-nm and 7-nm carbon nanotube field-effect transistor (CNFET) technologies. Our CNFET-OCL is designed to emulate the predictive 5-nm and 7-nm CNFET technologies presented in a published paper. We achieved this by performing a number of SPICE simulations based on an open-source CNFET SPICE model and making certain assumptions used in previous work. Each of our cell libraries includes two types of delay model (i.e., the composite current source and nonlinear delay model), each having 56 typical standard cells, which is sufficient to design various VLSI circuits. CNFET-OCL fully supports both logic synthesis and timing-driven place and route design in the Cadence design flow. Our experimental results demonstrate that CNFET-OCL can achieve performance levels comparable to those reported in previous studies on CNFETs. Consequently, CNFET-OCL can serve as an effective evaluation tool for the CNFET research community.
- 7 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 165335-165347
Open AccessCNFET-OCL: Open-Source Cell Libraries for Advanced CNFET Technologies
In this paper, we propose CNFET-OCL, the first open-source cell libraries for 5-nm and 7-nm carbon nanotube field-effect transistor (CNFET) technologies. Our CNFET-OCL is designed to emulate the predictive 5-nm and 7-nm CNFET technologies presented in a published paper. We achieved this by performing a number of SPICE simulations based on an open-source CNFET SPICE model and making certain assumptions used in previous work. Each of our cell libraries includes two types of delay model (i.e., the composite current source and nonlinear delay model), each having 56 typical standard cells, which is sufficient to design various VLSI circuits. CNFET-OCL fully supports both logic synthesis and timing-driven place and route design in the Cadence design flow. Our experimental results demonstrate that CNFET-OCL can achieve performance levels comparable to those reported in previous studies on CNFETs. Consequently, CNFET-OCL can serve as an effective evaluation tool for the CNFET research community.
- 1 December 2020
- Published by Institute of Electronics, Information and Communications Engineers (IEICE) in IEICE Transactions on Information and Systems
- Vol. E103.D (12) , 2590-2599
RPC: An Approach for Reducing Compulsory Misses in Packet Processing Cache
We propose a technique to reduce compulsory misses of packet processing cache (PPC), which largely affects both throughput and energy of core routers. Rather than prefetching data, our technique called response prediction cache (RPC) speculatively stores predicted data in PPC without additional access to the low-throughput and power-consuming memory (i.e., TCAM). RPC predicts the data related to a response flow at the arrival of the corresponding request flow, based on the request-response model of internet communications. Our experimental results with 11 real-network traces show that RPC can reduce the PPC miss rate by 13.4% in upstream and 47.6% in downstream on average when we suppose three-layer PPC. Moreover, we extend RPC to adaptive RPC (A-RPC) that selects the use of RPC in each direction within a core router for further improvement in PPC misses. Finally, we show that A-RPC can achieve 1.38x table-lookup throughput with 74% energy consumption per packet, when compared to conventional PPC.
- 1 December 2020
- Published by Institute of Electronics, Information and Communications Engineers (IEICE) in IEICE Transactions on Information and Systems
- Vol. E103.D (12) , 2590-2599
RPC: An Approach for Reducing Compulsory Misses in Packet Processing Cache
We propose a technique to reduce compulsory misses of packet processing cache (PPC), which largely affects both throughput and energy of core routers. Rather than prefetching data, our technique called response prediction cache (RPC) speculatively stores predicted data in PPC without additional access to the low-throughput and power-consuming memory (i.e., TCAM). RPC predicts the data related to a response flow at the arrival of the corresponding request flow, based on the request-response model of internet communications. Our experimental results with 11 real-network traces show that RPC can reduce the PPC miss rate by 13.4% in upstream and 47.6% in downstream on average when we suppose three-layer PPC. Moreover, we extend RPC to adaptive RPC (A-RPC) that selects the use of RPC in each direction within a core router for further improvement in PPC misses. Finally, we show that A-RPC can achieve 1.38x table-lookup throughput with 74% energy consumption per packet, when compared to conventional PPC.
- 16 July 2020
- Published by Wiley in International Journal of Communication Systems
- Vol. 33 (14)
Effective cache replacement policy for packet processing cache
Summary: Owing to the explosive increase in Internet traffic, increasing the throughput and energy efficiency of core routers is an important issue for future internet. Accessing the ternary content addressable memory (TCAM) in a core router can significantly degrade its throughput and energy efficiency owing to its low operating frequency and power consuming operations. To improve the throughput and energy‐efficiency of core routers, packet processing caches (PPCs) can be potentially used. PPCs can reduce the number of TCAM accesses by storing the TCAM lookup results in a fast and low‐energy cache memory and reusing them to process subsequent packets. To enhance the PPC performance, it is important to reduce the number of PPC misses. However, the PPC miss rate of the state‐of‐the‐art PPC is still high owing to its small capacity. In this study, we propose an effective cache replacement policy, called elevator cache (ELC) to improve the PPC miss rate. Our simulation results showed that the 8‐way ELC, which can be implementable at low hardware cost, reduced the number of PPC misses by 17.1% and 5.67% in comparison to the 4‐way least recently used (LRU) and the 4‐way ELC, respectively. Consequently, it was found that the 8‐way ELC can achieve a 1.21 times higher throughput (560 Gbps) while consuming 83% energy in comparison to conventional PPCs.
- 16 July 2020
- Published by Wiley in International Journal of Communication Systems
- Vol. 33 (14)
Effective cache replacement policy for packet processing cache
Summary: Owing to the explosive increase in Internet traffic, increasing the throughput and energy efficiency of core routers is an important issue for future internet. Accessing the ternary content addressable memory (TCAM) in a core router can significantly degrade its throughput and energy efficiency owing to its low operating frequency and power consuming operations. To improve the throughput and energy‐efficiency of core routers, packet processing caches (PPCs) can be potentially used. PPCs can reduce the number of TCAM accesses by storing the TCAM lookup results in a fast and low‐energy cache memory and reusing them to process subsequent packets. To enhance the PPC performance, it is important to reduce the number of PPC misses. However, the PPC miss rate of the state‐of‐the‐art PPC is still high owing to its small capacity. In this study, we propose an effective cache replacement policy, called elevator cache (ELC) to improve the PPC miss rate. Our simulation results showed that the 8‐way ELC, which can be implementable at low hardware cost, reduced the number of PPC misses by 17.1% and 5.67% in comparison to the 4‐way least recently used (LRU) and the 4‐way ELC, respectively. Consequently, it was found that the 8‐way ELC can achieve a 1.21 times higher throughput (560 Gbps) while consuming 83% energy in comparison to conventional PPCs.
- 25 February 2020
- Published by Springer Nature
- p. 338-352
Efficient Cache Architecture for Packet Processing in Internet Routers
- 25 February 2020
- Published by Springer Nature
- p. 338-352
Efficient Cache Architecture for Packet Processing in Internet Routers
- 4 October 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 69 (2) , 172-184
Footprint-Based DIMM Hotplug
Power-efficiency has become one of the most critical concerns for HPC as we continue to scale computational capabilities. A significant fraction of system power is spent on large main memories, mainly caused by the substantial amount of DIMM standby power needed. However, while necessary for some workloads, for many workloads large memory configurations are too rich, i.e., these workloads only make use of a fraction of the available memory, causing unnecessary power usage. This observation opens new opportunities for power reduction by powering DIMMs on and off depending on the current workload. In this paper, we propose footprint-based DIMM hotplug that enables a compute node to adjust the number of DIMMs that are powered on depending on the memory footprint of a running job. Our technique relies on two main subcomponents-memory footprint monitoring and DIMM management-which we both implement as part of an optimized page management system with small control overhead. Using Linux's memory hotplug capabilities, we implement our approach on a real system, and our results show that our proposed technique can save 50.3–51.7% of the DIMM standby energy and the CPU+DRAM energy of up to 1.48 Wh for various small-memory-footprint applications without loss of performance.
- 4 October 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 69 (2) , 172-184
Footprint-Based DIMM Hotplug
Power-efficiency has become one of the most critical concerns for HPC as we continue to scale computational capabilities. A significant fraction of system power is spent on large main memories, mainly caused by the substantial amount of DIMM standby power needed. However, while necessary for some workloads, for many workloads large memory configurations are too rich, i.e., these workloads only make use of a fraction of the available memory, causing unnecessary power usage. This observation opens new opportunities for power reduction by powering DIMMs on and off depending on the current workload. In this paper, we propose footprint-based DIMM hotplug that enables a compute node to adjust the number of DIMMs that are powered on depending on the memory footprint of a running job. Our technique relies on two main subcomponents-memory footprint monitoring and DIMM management-which we both implement as part of an optimized page management system with small control overhead. Using Linux's memory hotplug capabilities, we implement our approach on a real system, and our results show that our proposed technique can save 50.3–51.7% of the DIMM standby energy and the CPU+DRAM energy of up to 1.48 Wh for various small-memory-footprint applications without loss of performance.