Bo-Cheng Lai
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
Bo-Cheng Lai has a total of 229 co-authors. Here are the top co-authors by number of shared publications:
- Chen-Feng Wang
Institute of Electronics, National Chiao Tung University, Hsinchu 300, Taiwan
3 shared publications - Mark S. Neubauer
University of Illinois
3 shared publications - Santosh Parajuli
University of Illinois
3 shared publications - Shih-Chieh Hsu
University of Washington
3 shared publications - Szu-Chia Chen
School of Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung 807, Taiwan
3 shared publications - Alexander Tapper
Imperial College London
2 shared publications
Latest Publications
- Open Access
- 13 October 2026
- Published by American Physical Society (APS) in PRX Intelligence
- Vol. 1 (2) , 021201
Machine Learning on Heterogeneous, Edge, and Quantum Hardware for Particle Physics
The next generation of particle physics experiments will face an era of challenges in data acquisition, due to unprecedented data rates and volumes along with extreme environments and operational constraints. Harnessing these data for scientific discovery demands real-time inference and decision-making, intelligent data reduction, and efficient processing architectures beyond current capabilities. Crucial to the success of this experimental paradigm are several emerging technologies, such as artificial intelligence and machine learning, silicon microelectronics, and the advent of quantum algorithms and processing. Their intersection includes areas of research such as low-power and low-latency devices for edge computing, heterogeneous accelerator systems, reconfigurable hardware, codesign and synthesis strategies, readout for cryogenic or high-radiation environments, and analog computing. This white paper presents a community-driven vision to identify and prioritize research and development opportunities in hardware-based ML systems and corresponding physics applications, contributing toward a successful transition to the data frontier of fundamental science.
- 13 October 2026
- Published by American Physical Society (APS) in PRX Intelligence
- Vol. 1 (2) , 021201
Open AccessMachine Learning on Heterogeneous, Edge, and Quantum Hardware for Particle Physics
The next generation of particle physics experiments will face an era of challenges in data acquisition, due to unprecedented data rates and volumes along with extreme environments and operational constraints. Harnessing these data for scientific discovery demands real-time inference and decision-making, intelligent data reduction, and efficient processing architectures beyond current capabilities. Crucial to the success of this experimental paradigm are several emerging technologies, such as artificial intelligence and machine learning, silicon microelectronics, and the advent of quantum algorithms and processing. Their intersection includes areas of research such as low-power and low-latency devices for edge computing, heterogeneous accelerator systems, reconfigurable hardware, codesign and synthesis strategies, readout for cryogenic or high-radiation environments, and analog computing. This white paper presents a community-driven vision to identify and prioritize research and development opportunities in hardware-based ML systems and corresponding physics applications, contributing toward a successful transition to the data frontier of fundamental science.
- Top Cited PapersOpen Access
- 27 May 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Reconfigurable Technology and Systems
- Vol. 19 (2) , 1-35
hls4ml: A Flexible, Open Source Platform for Deep Learning Acceleration on Reconfigurable Hardware
We present hls4ml , a free and open source platform that translates machine learning (ML) models from modern deep learning frameworks into high-level synthesis (HLS) code that can be integrated into full designs for field-programmable gate arrays (FPGAs) or application-specific integrated circuits (ASICs). With its flexible and modular design, hls4ml supports a large number of deep learning frameworks and can target HLS compilers from several vendors, including Vitis HLS, Intel oneAPI and Catapult HLS. Together with a wider eco-system for software-hardware co-design, hls4ml has enabled the acceleration of ML inference in a wide range of commercial and scientific applications where low latency, resource usage, and power consumption are critical. In this article, we describe the structure and functionality of the hls4ml platform. The overarching design considerations for the generated HLS code are discussed, together with selected performance results.
- 27 May 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Reconfigurable Technology and Systems
- Vol. 19 (2) , 1-35
Top Cited PapersOpen Accesshls4ml: A Flexible, Open Source Platform for Deep Learning Acceleration on Reconfigurable Hardware
We present hls4ml , a free and open source platform that translates machine learning (ML) models from modern deep learning frameworks into high-level synthesis (HLS) code that can be integrated into full designs for field-programmable gate arrays (FPGAs) or application-specific integrated circuits (ASICs). With its flexible and modular design, hls4ml supports a large number of deep learning frameworks and can target HLS compilers from several vendors, including Vitis HLS, Intel oneAPI and Catapult HLS. Together with a wider eco-system for software-hardware co-design, hls4ml has enabled the acceleration of ML inference in a wide range of commercial and scientific applications where low latency, resource usage, and power consumption are critical. In this article, we describe the structure and functionality of the hls4ml platform. The overarching design considerations for the generated HLS code are discussed, together with selected performance results.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 33, 3347-3351
Switchable-Precision Universally Slimmable Networks
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 33, 3347-3351
Switchable-Precision Universally Slimmable Networks
- 1 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 25 (23) , 43256-43266
Design of a Low-Cost, High-Resolution Underwater Acoustic Imager With Integrated Deep Learning Enhancement
- 1 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 25 (23) , 43256-43266
Design of a Low-Cost, High-Resolution Underwater Acoustic Imager With Integrated Deep Learning Enhancement
- Open Access
- 27 February 2025
- Published by Association for Computing Machinery (ACM)
- p. 51
BAQET: BRAM-aware Quantization for Efficient Transformer Inference via Stream-based Architecture on an FPGA
FPGAs are a compelling substrate for supporting machine learning inference. Tools such as High-Level Synthesis and hls4ml can shorten the development cycle for deploying ML algorithms on FPGAs, but can struggle to handle the large on-chip storage needed for many of these models. In particular the high BRAM usage found in many of these flows can cause Place & Route failures during synthesis. In this paper we propose using a Simulated-Annealing based flow to perform BRAM-aware quantization. This approach trades off inference accuracy with BRAM usage, to provide a high-quality inference engine that still meets on-chip resource constraints. We demonstrate this flow for Transformer-based machine learning algorithms, which include Flash Attention in a Stream-based Dataflow architecture. Our system imposes minimal accuracy drops, yet can reduce BRAM usage by 20%-50%, and improve power efficiency by 264%-812% compared to existing Transformer-based accelerators on FPGAs
- 27 February 2025
- Published by Association for Computing Machinery (ACM)
- p. 51
Open AccessBAQET: BRAM-aware Quantization for Efficient Transformer Inference via Stream-based Architecture on an FPGA
FPGAs are a compelling substrate for supporting machine learning inference. Tools such as High-Level Synthesis and hls4ml can shorten the development cycle for deploying ML algorithms on FPGAs, but can struggle to handle the large on-chip storage needed for many of these models. In particular the high BRAM usage found in many of these flows can cause Place & Route failures during synthesis. In this paper we propose using a Simulated-Annealing based flow to perform BRAM-aware quantization. This approach trades off inference accuracy with BRAM usage, to provide a high-quality inference engine that still meets on-chip resource constraints. We demonstrate this flow for Transformer-based machine learning algorithms, which include Flash Attention in a Stream-based Dataflow architecture. Our system imposes minimal accuracy drops, yet can reduce BRAM usage by 20%-50%, and improve power efficiency by 264%-812% compared to existing Transformer-based accelerators on FPGAs
- Open Access
- 27 February 2025
- Published by Association for Computing Machinery (ACM)
- p. 47
HiGTR: High-Performance FPGA Implementation of Complete GNN-based Trajectory Reconstruction for HEP
Charged particle trajectory reconstruction is a critical task in high-energy physics (HEP), particularly for collision analysis in the Large Hadron Collider (LHC). In the LHC, the Level-1 Trigger (L1T) system must perform trajectory reconstruction with ultralow latency and very high throughput. Graph Neural Networks (GNNs)-based trajectory reconstruction on FPGAs has shown promising performance. However, the existing FPGA-based implementations are incomplete, supporting only the GNN processing stage and lacking the graph construction and track building parts of the task flow. In this paper, we present HiGTR, a high-performance FPGA implementation for complete GNN-based trajectory reconstruction, integrating graph construction, GNN processing, and track building. The existing GNN FPGA design is enhanced, achieving a 52.3% reduction in latency. The graph construction and track building algorithms are refined to significantly reduce computational complexity and enhance their suitability for FPGA implementation. The integrated HiGTR utilizes data streaming and multi-granularity pipelining with the flexibility to support varying data volumes. Implemented on an AMD Xilinx VU9P FPGA, HiGTR achieves a speedup of 65,204× compared to the previous software-based algorithm. In addition, HiGTR meets the stringent performance requirements of the L1T system in the HL-LHC upgrade project, including a throughput of 2.22 MHz and a latency of 4 microseconds. These results highlight its strong potential for practical deployment in the HL-LHC upgrade project.
- 27 February 2025
- Published by Association for Computing Machinery (ACM)
- p. 47
Open AccessHiGTR: High-Performance FPGA Implementation of Complete GNN-based Trajectory Reconstruction for HEP
Charged particle trajectory reconstruction is a critical task in high-energy physics (HEP), particularly for collision analysis in the Large Hadron Collider (LHC). In the LHC, the Level-1 Trigger (L1T) system must perform trajectory reconstruction with ultralow latency and very high throughput. Graph Neural Networks (GNNs)-based trajectory reconstruction on FPGAs has shown promising performance. However, the existing FPGA-based implementations are incomplete, supporting only the GNN processing stage and lacking the graph construction and track building parts of the task flow. In this paper, we present HiGTR, a high-performance FPGA implementation for complete GNN-based trajectory reconstruction, integrating graph construction, GNN processing, and track building. The existing GNN FPGA design is enhanced, achieving a 52.3% reduction in latency. The graph construction and track building algorithms are refined to significantly reduce computational complexity and enhance their suitability for FPGA implementation. The integrated HiGTR utilizes data streaming and multi-granularity pipelining with the flexibility to support varying data volumes. Implemented on an AMD Xilinx VU9P FPGA, HiGTR achieves a speedup of 65,204× compared to the previous software-based algorithm. In addition, HiGTR meets the stringent performance requirements of the L1T system in the HL-LHC upgrade project, including a throughput of 2.22 MHz and a latency of 4 microseconds. These results highlight its strong potential for practical deployment in the HL-LHC upgrade project.
- 28 April 2023
- Published by Elsevier in Engineering Structures
- Vol. 288, 116186
Development and seismic performance evaluation of New high strength reinforced concrete column and steel beam (New-RCS) joint
- 28 April 2023
- Published by Elsevier in Engineering Structures
- Vol. 288, 116186
Development and seismic performance evaluation of New high strength reinforced concrete column and steel beam (New-RCS) joint
- 27 January 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE/ACM Transactions on Computational Biology and Bioinformatics
- Vol. 20 (3) , 2210-2222
A Bin-Based Indexing for Scalable Range Join on Genomic Data
Range-join is an operation for finding overlaps in interval-form genomic data. Range-join is widely used in various genome analysis processes such as annotation, filtering and comparison of variants in whole-genome and exome analysis pipelines. The quadratic complexity of current algorithms with sheer data volume has surged the design challenges. Existing tools have limitations on algorithm efficiency, parallelism, scalability and memory consumption. This paper proposes BIndex, a novel bin-based indexing algorithm and its distributed implementation to attain high throughput range-join processing. BIndex features near-constant search complexity while the inherently parallel data structure facilitates exploitation of parallel computing architectures. Balanced partitioning of dataset further enables scalability on distributed frameworks. The implementation on Message Passing Interface shows upto 933.5x speedup in comparison to state-of-the-art tools. Parallel nature of BIndex further enables GPU-based acceleration with 3.72x speedup than CPU implementations. The add-in modules for Apache Spark provides upto 4.65x speedup than the previously best available tool. BIndex supports wide variety of input and output formats prevalent in bioinformatics community and the algorithm is easily extendable to streaming data in recent Big Data solutions. Furthermore, the index data structure is memory-efficient and consumes upto two orders-of-magnitude lesser RAM, while having no adverse effect on speedup.
- 27 January 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE/ACM Transactions on Computational Biology and Bioinformatics
- Vol. 20 (3) , 2210-2222
A Bin-Based Indexing for Scalable Range Join on Genomic Data
Range-join is an operation for finding overlaps in interval-form genomic data. Range-join is widely used in various genome analysis processes such as annotation, filtering and comparison of variants in whole-genome and exome analysis pipelines. The quadratic complexity of current algorithms with sheer data volume has surged the design challenges. Existing tools have limitations on algorithm efficiency, parallelism, scalability and memory consumption. This paper proposes BIndex, a novel bin-based indexing algorithm and its distributed implementation to attain high throughput range-join processing. BIndex features near-constant search complexity while the inherently parallel data structure facilitates exploitation of parallel computing architectures. Balanced partitioning of dataset further enables scalability on distributed frameworks. The implementation on Message Passing Interface shows upto 933.5x speedup in comparison to state-of-the-art tools. Parallel nature of BIndex further enables GPU-based acceleration with 3.72x speedup than CPU implementations. The add-in modules for Apache Spark provides upto 4.65x speedup than the previously best available tool. BIndex supports wide variety of input and output formats prevalent in bioinformatics community and the algorithm is easily extendable to streaming data in recent Big Data solutions. Furthermore, the index data structure is memory-efficient and consumes upto two orders-of-magnitude lesser RAM, while having no adverse effect on speedup.
- 4 May 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Emerging and Selected Topics in Circuits and Systems
- Vol. 12 (2) , 486-499
DSIM: Distributed Sequence Matching on Near-DRAM Accelerator for Genome Assembly
Matching nucleic acid sequences has long become the performance bottleneck in genome assembly which aims to connect enormous partial genome reads without prior knowledge of the reference sequence. The intensive and random data accesses of querying sequences using the widely adopted FM-Index data structure have caused in-efficient usage and long runtime of the memory system. Existing software FM-Index tools are limited on algorithmic inefficiency and poor processing parallelism. Solutions on GPU, FPGA and ASIC focus mainly on computational acceleration while still bottlenecked at the memory-bound nature of querying FM-Index. This paper proposes DSIM, a scalable FM-Index querying on near-DRAM accelerators. DSIM supports highly parallel multi-step query processing by distributing partial FM-Index table to different DRAM chips. Each genome sequence is partitioned into shorter queries and dispatched to the corresponding DRAM chip for string lookup. The optimized data layout and execution control on DRAM enables high row-data reuse and minimizes CPU-DRAM data transfers. The light-weight mapping scheme on the host CPU facilitates effective query distribution to DRAM chips and further supports scalability to multiple DIMMs (Dual-Inline Memory Modules). An in-DRAM arbiter is implemented to control the intra-chip data processing without affecting the external memory controller and DDR protocol. Experiments on 128-chip DRAM system showed that DSIM achieves up to 231 $\times$ and 8.9 $\times$ overall speedup compared to the software FM-Index tool and the state-of-the-art near-DRAM solution respectively.
- 4 May 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Emerging and Selected Topics in Circuits and Systems
- Vol. 12 (2) , 486-499
DSIM: Distributed Sequence Matching on Near-DRAM Accelerator for Genome Assembly
Matching nucleic acid sequences has long become the performance bottleneck in genome assembly which aims to connect enormous partial genome reads without prior knowledge of the reference sequence. The intensive and random data accesses of querying sequences using the widely adopted FM-Index data structure have caused in-efficient usage and long runtime of the memory system. Existing software FM-Index tools are limited on algorithmic inefficiency and poor processing parallelism. Solutions on GPU, FPGA and ASIC focus mainly on computational acceleration while still bottlenecked at the memory-bound nature of querying FM-Index. This paper proposes DSIM, a scalable FM-Index querying on near-DRAM accelerators. DSIM supports highly parallel multi-step query processing by distributing partial FM-Index table to different DRAM chips. Each genome sequence is partitioned into shorter queries and dispatched to the corresponding DRAM chip for string lookup. The optimized data layout and execution control on DRAM enables high row-data reuse and minimizes CPU-DRAM data transfers. The light-weight mapping scheme on the host CPU facilitates effective query distribution to DRAM chips and further supports scalability to multiple DIMMs (Dual-Inline Memory Modules). An in-DRAM arbiter is implemented to control the intra-chip data processing without affecting the external memory controller and DDR protocol. Experiments on 128-chip DRAM system showed that DSIM achieves up to 231 $\times$ and 8.9 $\times$ overall speedup compared to the software FM-Index tool and the state-of-the-art near-DRAM solution respectively.
- Open Access
- 21 August 2021
- Published by MDPI AG in Journal of Personalized Medicine
- Vol. 11 (8) , 819
The Impact of the Synergistic Effect of Temperature and Air Pollutants on Chronic Lung Diseases in Subtropical Taiwan
Previous studies have suggested an association between air pollution and lung disease. However, few studies have explored the relationship between chronic lung diseases classified by lung function and environmental parameters. This study aimed to comprehensively investigate the relationship between chronic lung diseases, air pollution, meteorological factors, and anthropometric indices. We conducted a cross-sectional study using the Taiwan Biobank and the Taiwan Air Quality Monitoring Database. A total of 2889 participants were included. We found a V/U-shaped relationship between temperature and air pollutants, with significant effects at both high and low temperatures. In addition, at lower temperatures (x) (aOR:1.02/ppm, 95% CI 1.00–1.05; aOR:1.06/ppm, 95% CI 1.04–1.08), and sulfur dioxide (SO2) (aOR:1.29/ppm, 95% CI 1.01–1.65; aOR:1.77/ppm, 95% CI 1.36–2.30) were associated with restrictive and mixed lung diseases, respectively. Exposure to CO, NO, NO2, NOx and SO2 significantly affected obstructive and mixed lung disease in southern Taiwan. In conclusion, temperature and air pollution should be considered together when evaluating the impact on chronic lung diseases.
- 21 August 2021
- Published by MDPI AG in Journal of Personalized Medicine
- Vol. 11 (8) , 819
Open AccessThe Impact of the Synergistic Effect of Temperature and Air Pollutants on Chronic Lung Diseases in Subtropical Taiwan
Previous studies have suggested an association between air pollution and lung disease. However, few studies have explored the relationship between chronic lung diseases classified by lung function and environmental parameters. This study aimed to comprehensively investigate the relationship between chronic lung diseases, air pollution, meteorological factors, and anthropometric indices. We conducted a cross-sectional study using the Taiwan Biobank and the Taiwan Air Quality Monitoring Database. A total of 2889 participants were included. We found a V/U-shaped relationship between temperature and air pollutants, with significant effects at both high and low temperatures. In addition, at lower temperatures (x) (aOR:1.02/ppm, 95% CI 1.00–1.05; aOR:1.06/ppm, 95% CI 1.04–1.08), and sulfur dioxide (SO2) (aOR:1.29/ppm, 95% CI 1.01–1.65; aOR:1.77/ppm, 95% CI 1.36–2.30) were associated with restrictive and mixed lung diseases, respectively. Exposure to CO, NO, NO2, NOx and SO2 significantly affected obstructive and mixed lung disease in southern Taiwan. In conclusion, temperature and air pollution should be considered together when evaluating the impact on chronic lung diseases.