Kazutomo Yoshii
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Kazutomo Yoshii has a total of 93 co-authors. Here are the top co-authors by number of shared publications:
- Antonino Miceli
Argonne National Laboratory, Lemont, IL, USA
6 shared publications - Franck Cappello
Argonne National Laboratory, Lemont, USA
4 shared publications - Sheng Di
Argonne National Laboratory, Lemont, IL, USA
3 shared publications - Dingwen Tao
Chinese Academy of Sciences
2 shared publications - Hanqi Guo
Ohio State University, Columbus, OH, USA
2 shared publications - Ian Foster
Argonne National Laboratory, Argonne, IL, USA
2 shared publications
Latest Publications
- Open Access
- 15 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 249-258
Design and Implementation of a Custom Hardware Accelerator for SZx Compression in Chipyard
- 15 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 249-258
Open AccessDesign and Implementation of a Custom Hardware Accelerator for SZx Compression in Chipyard
- Open Access
- 15 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 2347-2353
Adapting scientific streaming inference workflows for a deterministic tensor processing unit
- 15 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 2347-2353
Open AccessAdapting scientific streaming inference workflows for a deterministic tensor processing unit
- Open Access
- 15 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 1736-1742
Assessing a RISC-V Accelerator for Cross-Section Lookup in Chipyard
- 15 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 1736-1742
Open AccessAssessing a RISC-V Accelerator for Cross-Section Lookup in Chipyard
- Open Access
- 15 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 1966-1979
What to Support When You’re Compressing
- 15 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 1966-1979
Open AccessWhat to Support When You’re Compressing
- 1 October 2025
- Published by IOP Publishing in Journal of Instrumentation
- Vol. 20 (10) , P10027
A 28 nm multiply-accumulate ASIC architecture for on-chip data compression in MHz frame rate X-ray and electron pixel detectors
Modern X-ray detector systems urgently require compact, efficient, and fast data compression schemes to handle the transmission of big data from pixel arrays, enabling frame rates in the MHz regime. In this work, a data compression ASIC that implements a streaming fixed-length lossy compression scheme is introduced and analyzed, proving the feasibility and benefits of on-chip compression. The compression scheme utilizes a vector matrix product logic, which performs a number of floating-point multiplications, additions, and accumulations. The logic is verified, synthesized, and shown to fit in the area resource available for the X-ray detector under study, which comprises 192 × 168 pixels each of 12-bit width, and having a total area of 20 mm× 20 mm, about 2 mm× 20 mm of which are available for the digital logic. Several system architectures, precisions, and compression ratios ranging from 100 to 250 were analyzed to pave the way for on-chip fixed-length compression (e.g., principal component analysis, singular value decomposition) and data reduction (e.g., azimuthal integration) for X-ray and electron detectors.
- 1 October 2025
- Published by IOP Publishing in Journal of Instrumentation
- Vol. 20 (10) , P10027
A 28 nm multiply-accumulate ASIC architecture for on-chip data compression in MHz frame rate X-ray and electron pixel detectors
Modern X-ray detector systems urgently require compact, efficient, and fast data compression schemes to handle the transmission of big data from pixel arrays, enabling frame rates in the MHz regime. In this work, a data compression ASIC that implements a streaming fixed-length lossy compression scheme is introduced and analyzed, proving the feasibility and benefits of on-chip compression. The compression scheme utilizes a vector matrix product logic, which performs a number of floating-point multiplications, additions, and accumulations. The logic is verified, synthesized, and shown to fit in the area resource available for the X-ray detector under study, which comprises 192 × 168 pixels each of 12-bit width, and having a total area of 20 mm× 20 mm, about 2 mm× 20 mm of which are available for the digital logic. Several system architectures, precisions, and compression ratios ranging from 100 to 250 were analyzed to pave the way for on-chip fixed-length compression (e.g., principal component analysis, singular value decomposition) and data reduction (e.g., azimuthal integration) for X-ray and electron detectors.
- Open Access
- 22 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 110-114
Directing PDES and Surrogate Models in Loosely Coupled Hybrid Simulations
- 22 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 110-114
Open AccessDirecting PDES and Surrogate Models in Loosely Coupled Hybrid Simulations
- Top Cited PapersOpen Access
- 11 June 2025
- Published by Association for Computing Machinery (ACM) in ACM Computing Surveys
- Vol. 57 (11) , 1-38
A Survey on Error-Bounded Lossy Compression for Scientific Datasets
Error-bounded lossy compression has been effective in significantly reducing the data storage/transfer burden while preserving the reconstructed data fidelity very well. Many error-bounded lossy compressors have been developed for a wide range of parallel and distributed use cases for years. They are designed with distinct compression models and principles, such that each of them features particular pros and cons. In this article, we provide a comprehensive survey of emerging error-bounded lossy compression techniques. The key contribution is fourfold. (1) We summarize a novel taxonomy of lossy compression into six classic models. (2) We provide a comprehensive survey of 10 commonly used compression components/modules. (3) We summarized pros and cons of 47 state-of-the-art lossy compressors and present how state-of-the-art compressors are designed based on different compression techniques. (4) We discuss how customized compressors are designed for specific scientific applications and use-cases. We believe this survey is useful to multiple communities including scientific applications, high-performance computing, lossy compression, and big data.
- 11 June 2025
- Published by Association for Computing Machinery (ACM) in ACM Computing Surveys
- Vol. 57 (11) , 1-38
Top Cited PapersOpen AccessA Survey on Error-Bounded Lossy Compression for Scientific Datasets
Error-bounded lossy compression has been effective in significantly reducing the data storage/transfer burden while preserving the reconstructed data fidelity very well. Many error-bounded lossy compressors have been developed for a wide range of parallel and distributed use cases for years. They are designed with distinct compression models and principles, such that each of them features particular pros and cons. In this article, we provide a comprehensive survey of emerging error-bounded lossy compression techniques. The key contribution is fourfold. (1) We summarize a novel taxonomy of lossy compression into six classic models. (2) We provide a comprehensive survey of 10 commonly used compression components/modules. (3) We summarized pros and cons of 47 state-of-the-art lossy compressors and present how state-of-the-art compressors are designed based on different compression techniques. (4) We discuss how customized compressors are designed for specific scientific applications and use-cases. We believe this survey is useful to multiple communities including scientific applications, high-performance computing, lossy compression, and big data.
- Open Access
- 19 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 10-17
BraggHLS: High-Level Synthesis for Low-Latency Deep Neural Networks for Experimental Science
In many experiment-driven scientific domains, such as high-energy physics, material science, and cosmology, high data rate experiments impose hard constraints on data acquisition systems: collected data must either be indiscriminately stored for post-processing and analysis, thereby necessitating large storage capacity, or accurately filtered in real-time, thereby necessitating low-latency processing. Deep neural networks, effective in other filtering tasks, have not been widely employed in such data acquisition systems, due to design and deployment difficulties. We present an open source, lightweight, compiler framework, without any proprietary dependencies, BraggHLS, based on high-level synthesis techniques, for translating high-level representations of deep neural networks to low-level representations, suitable for deployment to near-sensor devices such as field-programmable gate arrays. We evaluate BraggHLS on various workloads and present a case-study implementation of a deep neural network for Bragg peak detection in the context of high-energy diffraction microscopy. We show BraggHLS is able to produce an implementation of the network with a throughput of 4.8 µs/sample, which is approximately a 4 × improvement over the existing implementation.
- 19 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 10-17
Open AccessBraggHLS: High-Level Synthesis for Low-Latency Deep Neural Networks for Experimental Science
In many experiment-driven scientific domains, such as high-energy physics, material science, and cosmology, high data rate experiments impose hard constraints on data acquisition systems: collected data must either be indiscriminately stored for post-processing and analysis, thereby necessitating large storage capacity, or accurately filtered in real-time, thereby necessitating low-latency processing. Deep neural networks, effective in other filtering tasks, have not been widely employed in such data acquisition systems, due to design and deployment difficulties. We present an open source, lightweight, compiler framework, without any proprietary dependencies, BraggHLS, based on high-level synthesis techniques, for translating high-level representations of deep neural networks to low-level representations, suitable for deployment to near-sensor devices such as field-programmable gate arrays. We evaluate BraggHLS on various workloads and present a case-study implementation of a deep neural network for Bragg peak detection in the context of high-energy diffraction microscopy. We show BraggHLS is able to produce an implementation of the network with a throughput of 4.8 µs/sample, which is approximately a 4 × improvement over the existing implementation.
- Open Access
- 12 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 1274-1278
Hardware Specialization: Estimating Monte Carlo Cross-Section Lookup Kernel Performance and Area
Hardware specialization is one of the promising directions in the post-Moore era. It is imperative to understand how hardware specialization paradigms can benefit HPC. An essential question revolves around estimating the theoretical performance of an optimally specialized architecture without requiring extensive hardware development expertise and efforts. Focusing on the Monte Carlo cross-section lookup kernel, known for its notably low resource utilization, we develop a workflow to simulate a specialized architecture’s timing and estimate resource usage to answer these questions, leveraging open-source hardware tools. We implement building blocks of the kernel pipeline in the Chisel construction language and generate Verilog codes for resource estimation. Our late-breaking results show that the kernel latency is 46 cycles per lookup while the optimized CPU code takes 680 cycles, and a potential 15k pipeline copies within a 698 mm2 die, reflective of the Intel Xeon Platinum 8180 dimensions.
- 12 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 1274-1278
Open AccessHardware Specialization: Estimating Monte Carlo Cross-Section Lookup Kernel Performance and Area
Hardware specialization is one of the promising directions in the post-Moore era. It is imperative to understand how hardware specialization paradigms can benefit HPC. An essential question revolves around estimating the theoretical performance of an optimally specialized architecture without requiring extensive hardware development expertise and efforts. Focusing on the Monte Carlo cross-section lookup kernel, known for its notably low resource utilization, we develop a workflow to simulate a specialized architecture’s timing and estimate resource usage to answer these questions, leveraging open-source hardware tools. We implement building blocks of the kernel pipeline in the Chisel construction language and generate Verilog codes for resource estimation. Our late-breaking results show that the kernel latency is 46 cycles per lookup while the optimized CPU code takes 680 cycles, and a potential 15k pipeline copies within a 698 mm2 die, reflective of the Intel Xeon Platinum 8180 dimensions.
- Open Access
- 12 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 289-297
Streaming Hardware Compressor Generator Framework
The interest in and strong demand for application-specific accelerators in computing and sensor data processing are rising. Simultaneously, data movement bottlenecks are increasingly becoming a significant limiting factor for these accelerators. Integrating an extremely resource-efficient, ultra-low-latency compressor block into their data path or pipeline could solve or mitigate data movement bottlenecks and enhance the performance of these accelerators. However, workflows for hardware compressor architecture exploration are little studied. We introduce a generator framework for designing, verifying, and estimating resources in streaming hardware compressor architectures to fill the gap. This framework assists users in exploring different compressor architectures with different compressor building blocks, evaluating their characteristics (latency, throughput, gate counts), and generating RTL code for integrating them into custom accelerator designs. Our motivation is to bridge the gap between software and hardware experts through this proposed framework as a co-design tool.
- 12 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 289-297
Open AccessStreaming Hardware Compressor Generator Framework
The interest in and strong demand for application-specific accelerators in computing and sensor data processing are rising. Simultaneously, data movement bottlenecks are increasingly becoming a significant limiting factor for these accelerators. Integrating an extremely resource-efficient, ultra-low-latency compressor block into their data path or pipeline could solve or mitigate data movement bottlenecks and enhance the performance of these accelerators. However, workflows for hardware compressor architecture exploration are little studied. We introduce a generator framework for designing, verifying, and estimating resources in streaming hardware compressor architectures to fill the gap. This framework assists users in exploring different compressor architectures with different compressor building blocks, evaluating their characteristics (latency, throughput, gate counts), and generating RTL code for integrating them into custom accelerator designs. Our motivation is to bridge the gap between software and hardware experts through this proposed framework as a co-design tool.