Seungkyu Choi

Contact (email): seungkc at khu dot ac dot kr

[Lab Webpage]   [Google Scholar]   [Linked-In]

Experience

• Mar. 2023 - Present Assistant Professor in the department of EE at Kyung Hee University, South Korea

• Mar. 2022 - Feb. 2023 Postdoctoral Associate at University of Michigan, Ann Arbor, MI, USA (PI: Prof. Hun-Seok Kim)

• Mar. 2016 - Feb. 2022 Graduate Student Researcher at KAIST, Daejeon, South Korea
  - Served as a student representative in MVLSI Lab. (2020) and a teaching assistant for Digital System / Digital Integrated Circuits (2017-2020)

• Jun. 2015 - Aug. 2015 Summer Internship at SK Hynix, Icheon, South Korea

Education

• Mar. 2018 ~ Feb. 2022

Ph.D. in School of Electrical Engineering, KAIST, Daejeon, South Korea

• Mar. 2016 ~ Feb. 2018

M.S. in School of Electrical Engineering, KAIST, Daejeon, South Korea

• Feb. 2011 ~ Feb. 2016

B.S. in School of Electrical Engineering, KAIST, Daejeon, South Korea (Cum Laude)

• Mar. 2009 ~ Feb. 2011

High school diploma in D(T)aegu Science High School  (completed in two years)

Publications

International Conference Papers

• ParaBase: A Configurable Parallel Baseband Processor for Ultra-High-Speed Inter-Satellite Optical Communications

Seungkyu Choi, Huanshihong Deng, Kuan-Yu Chen, Yufan Yue, David Blaauw, and Hun-Seok Kim

IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED), 2024 (Accepted)

• MERSIT: A Hardware-Efficient 8-bit Data Format with Enhanced Post-Training Quantization DNN Accuracy

Nguyen-Dong Ho, Gyujun Jeong, Cheol-Min Kang, Seungkyu Choi* and Ik Joon Chang*

ACM/IEEE Design Automation Conference (DAC), 2024 (Accepted)

• SONA: An Accelerator for Transform-Domain Neural Networks with Sparse-Orthogonal Weights

Pierre Abillama, Zichen Fan, Yu Chen, Hyochan An, Qirui Zhang, Seungkyu Choi, David Blaauw, Dennis Sylvester, and Hun-Seok Kim

IEEE International Conference on Application-specific Systems, Architectures, and Processors (ASAP), 2023   (Best Paper Nominee)

• Algorithm/Architecture Co-Design for Energy-Efficient Acceleration of Multi-Task DNN

Jaekang Shin, Seungkyu Choi, Jongwoo Ra, and Lee-Sup Kim

ACM/IEEE Design Automation Conference (DAC), 2022  [DOI]

• A Convergence Monitoring Method for DNN Training of On-device Task Adaptation

Seungkyu Choi, Jaekang Shin, and Lee-Sup Kim

IEEE/ACM International Conference on Computer-Aided Design (ICCAD), 2021 [DOI]

• A Pragmatic Approach to On-device Incremental Learning System with Selective Weight Updates

Jaekang Shin, Seungkyu Choi, Yeongjae Choi, and Lee-Sup Kim

ACM/IEEE Design Automation Conference (DAC), 2020 [DOI]

• A 47.4µJ/epoch Trainable Deep Convolutional Neural Network Accelerator for In-Situ Personalization on Smart Devices

Seungkyu Choi, Jaehyeong Sim, Myeonggu Kang, Yeongjae Choi, Hyeonuk Kim, and Lee-Sup Kim

IEEE Asian Solid-State Circuits Conference (A-SSCC), 2019 [DOI]

• Compressing Sparse Ternary Weight Convolutional Neural Networks for Efficient Hardware Acceleration

Hyeonwook Wi, Hyeonuk Kim, Seungkyu Choi, and Lee-Sup Kim

IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED), 2019 [DOI]

• An Optimized Design Technique of Low-bit Neural Network Training for Personalization on IoT Devices

Seungkyu Choi, Jaekang Shin, Yeongjae Choi, and Lee-Sup Kim

ACM/IEEE Design Automation Conference (DAC), 2019 [DOI]

• TrainWare: A Memory Optimized Weight Update Architecture for On-Device Convolutional Neural Network Training

Seungkyu Choi, Jaehyeong Sim, Myeonggu Kang, and Lee-Sup Kim

IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED), 2018 [DOI]

• SENIN: An Energy-efficient Sparse Neuromorphic System with On-chip Learning

Myung-Hoon Choi, Seungkyu Choi, Jaehyeong Sim, and Lee-Sup Kim

IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED), 2017 [DOI]

International Journal Papers

• Accelerating On-Device DNN Training Workloads via Runtime Convergence Monitor

Seungkyu Choi, Jaekang Shin, and Lee-Sup Kim

IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, May 2023 [DOI]

• Energy-Efficient CNN Personalized Training by Adaptive Data Reformation

Youngbeom Jung, Hyeonuk Kim, Seungkyu Choi, Jaekang Shin, and Lee-Sup Kim

IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, Jan. 2023 [DOI]

• A Deep Neural Network Training Architecture with Inference-aware Heterogeneous Data-type

Seungkyu Choi, Jaekang Shin, and Lee-Sup Kim

IEEE Transactions on Computers, May 2022 [DOI]

• Rare Computing: Removing Redundant Multiplications from Sparse and Repetitive Data in Deep Neural Networks

Kangkyu Park, Seungkyu Choi, Yeongjae Choi, and Lee-Sup Kim

IEEE Transactions on Computers, Apr. 2022 [DOI]

• An Energy-Efficient Deep Convolutional Neural Network Training Accelerator for In Situ Personalization on Smart Devices

Seungkyu Choi, Jaehyeong Sim, Myeonggu Kang, Yeongjae Choi, Hyeonuk Kim, and Lee-Sup Kim

IEEE Journal of Solid-State Circuits, Oct. 2020 [DOI]

• Energy-Efficient Design of Processing Element for Convolutional Neural Network

Yeongjae Choi, Dongmyung Bae, Jaehyeong Sim, Seungkyu Choi, Minhye Kim, and Lee-Sup Kim

IEEE Transactions on Circuits and Systems II, Nov. 2017 [DOI]

Patents

International 

• Method and Apparatus with Neural Network Compression (US-17892481)

• Method and Device for Encoding (US-17401453)

• Method and Apparatus with Incremental Learning Model (US-17089764; CN-202110055314.1) 

• Method and Apparatus with Neural Network Data Quantizing  (US-15931362; EP-20183579; CN-202011086791.6) 

Domestic 

• Processing Unit for Processing Multi-Tasks (KR-10-2021-0143670) 

• Method and Apparatus for Compressing a Neural Network (KR-10-2021-0143629) 

• Exponent Encoded Half-float Representation for Enabling Brick-level Fixed-point MAC (KR-10-2021-0028929) 

• Method of Binary Mask  (KR-10-2020-0041638) 

• Local Maximum Quantization for Convolutional Neural Network Training  (KR-10-2019-0126298) 

Honors/Awards

• IDEC Congress Chip Design Contest 2019 Best Poster Award

Main recipient, "A Deep Convolutional Neural Network Training Processor for Personalization on Smart Devices"

• KAIST Bachelor's Degree Cum Laude

• Korea Government Full Scholarship (KAIST)

Academic Services

Reviewer

• IEEE Transactions on Circuits and Systems I (TCAS-I):  2019, 2021, 2022, 2023

• IEEE Transactions on Circuits and Systems II (TCAS-II):  2022

• IEEE Open Journal of Circuits and Systems (OJCAS):  2020

• IEEE Open Journal of the Solid-State Circuits Society (OJSSCS):  2021

• IEEE Transactions on Emerging Topics in Computing (TETC): 2021, 2023

• IEEE Transactions on Magnetics: 2022