Research Interests

---

• High-performance computing applications such as deep learning.
• Scalability on heterogeneous clusters with accelerators such as GPUs and FPGAs.
• Utilizing storage devices (HDDs, NVMe SSDs, ...) to overcome memory capacity.

Education

---

• Mar. 2017 - Present

  Pursuing Ph.D. in Computer Science and Engineering

  Seoul National University

  Advisor: Prof. Jaejin Lee

• Mar. 2013 - Feb. 2017

  B.S. in Computer Science and Engineering

  Seoul National University

  Summa cum laude (Overall GPA: 4.07 / 4.30, Major GPA: 4.21 / 4.30)

• Mar. 2010 - Feb. 2013

  Korea Science Academy

  Busan, Korea

Publications

---

Conferences

• TCCL: Discovering Better Communication Paths for PCIe GPU Clusters.

  Heehoon Kim, Junyeol Ryu, and Jaejin Lee

  ASPLOS 2024: Proceedings of the 29th International Conference on Architectural Support for Programming Languages and Operating Systems, Vol. 3, pp. 999-1015, San Diego, USA, April 2024. DOI: 10.1145/3620666.3651362

• SnuQS: Scaling Quantum Circuit Simulation using Storage Devices.

  Daeyoung Park, Heehoon Kim, Jinpyo Kim, Taehyun Kim, and Jaejin Lee.

  ICS '22: Proceedings of the 36th ACM International Conference on Supercomputing, No. 6, pp. 1-13, Online, June 2022. DOI: 10.1145/3524059.3532375

• SOFF: An OpenCL High-Level Synthesis Framework for FPGAs.

  Gangwon Jo, Heehoon Kim, Jeesoo Lee, and Jaejin Lee.

  ISCA '20: Proceedings of the 47th Annual International Symposium on Computer Architecture, pp. 295-308, Online, June 2020. DOI: 10.1109/ISCA45697.2020.00034

• Performance Analysis of CNN Frameworks for GPUs.

  Heehoon Kim, Hyoungwook Nam, Wookeun Jung, and Jaejin Lee.

  ISPASS '17: Proceedings of 2017 IEEE International Symposium on Performance Analysis of Systems and Software, pp. 55-64, Santa Rosa, California, USA, April 2017. DOI: 10.1109/ISPASS.2017.7975270

Papers in Korean

• Implementation and Evaluation of a DMA Controller for PCIe-based FPGA Boards.

  Heehoon Kim, and Jaejin Lee.

  Journal of KIISE, 48(2), 141-146, February 2021. DOI: 10.5626/JOK.2021.48.2.141

  Best paper award. KCC '20: Proceedings of 2020 KIISE Korea Computer Congress, 1192-1194, July 2020. RISS: 107047457

• Performance Analysis of FPGA Off-Chip Memory.

  Heehoon Kim, and Jaejin Lee.

  KCC '19: Proceedings of 2019 KIISE Korea Computer Congress, 39-41, June 2019. RISS: 106329771

• Scalability of Deep Neural Network training on multi-GPU.

  Heehoon Kim, Wookeun Jung, and Jaejin Lee.

  KCC '17: Proceedings of 2017 KIISE Korea Computer Congress, 1474-1476, June 2017. RISS: 103298318

• Universal Heterogeneous Programming Environment.

  Jungho Park, Gangwon Jo, Wookeun Jung, Hyungmo Kim, Jaehoon Jung, Thanh Tuan Dao, Jiyoung Park, Jungwook Kim, Heehoon Kim, Youngdong Do, PyeongSeok Oh, and Jaejin Lee.

  Communications of KIISE, Vol. 35, No. 10, pp. 18-31, 2017. RISS: 103625426

• Hardware and Software Support for Deep Learning.

  Wookeun Jung, Jungwook Kim, Thanh Tuan Dao, Jungho Park, Jiyoung Park, Jaeho Shin, Jaehoon Jung, Gangwon Jo, Heehoon Kim, Hyoungwook Nam, and Jaejin Lee.

  Communications of KIISE, Vol. 34, No. 9, pp. 10-20, 2016. RISS: 102075344

Professional Activities

---

• Artifact Evaluation Committee

  PPoPP 2020

• Student Volunteers

  PPoPP 2022

  PPoPP 2021

  PACT 2021

Experiences

---

• Lecturer

  2024 Accelerator Programming Winter School

  Moreh-Vietnam GPU Engineer Training Program (Winter 2024)

  2023 Accelerator Programming Summer School

• Teaching Assistant @ Seoul National University

  Field Applications of Data Science (Fall 2024, M3239.003900)

  Bootcamp for Graduate School of Data Science (Winter 2021)

  Scalable High-Performance Computing (Fall 2020, M3239.003800)

  Multicore Computing (Spring 2020, 4190.414A)

  Logic Design (Spring 2018, M1522.000700)

  Multicore Computing (Fall 2017, 4190.414A)

  Programming Practice (Fall 2016, 4190.103A)

• Teaching Assistant @ Samsung

  System Architect/Expert Course 2023 - Parallel Programming

  System Architect/Expert Course 2022 - Parallel Programming

  NPU Expert Course 2021 - Deep Learning Accelerator Architectures

  NPU Expert Course 2020 - Deep Learning Models, Frameworks, and Programming

• Teaching Assistant @ Accelerator Programming School

  2023 Accelerator Programming Winter School

  2018 Accelerator Programming Summer School

  2018 Accelerator Programming Winter School

  2017 Accelerator Programming Summer School

  2017 Accelerator Programming Winter School

• Mar. 2013 - Feb. 2017

  SNUPS (Problem Solving Club of SNU)

  Club president at 2014

Honors & Awards

---

• Large Language Model

  1st place, Samsung Computer Engineering Challenge 2023 (Report)

  Fastest inference on HellaSwag with LLaMA-30B, in a server with four NVIDIA Tesla V100 GPUs

• Near-Memory Computing

  1st place, Open Innovation Contest for AXDIMM Technology by Samsung (Award: $50,000, Winners, Report)

  Synergistic Approach for Systems with AXDIMMs, GPUs, and NVMe Devices

• Supercomputing

  Grand prize, 2017 Korea Supercomputing Challenge (Photo)

• Competitive Programming (Handle: @csehydrogen)

  6th Place, 2016 ACM-ICPC Asia Daejeon Regional Contest

  3rd Place, 2016 Samsung Collegiate Programming Cup

  3rd Place, LG CNS Code Monster 2016

  6th Place, 2015 ACM-ICPC Asia Daejeon Regional Contest

  3rd Place, 2015 Samsung Collegiate Programming Cup

  4th Place, 2014 ACM-ICPC Asia Daejeon Regional Contest

  7th Place, 2013 ACM-ICPC Asia Danang Regional Contest

  10th Place, 2013 ACM-ICPC Asia Daejeon Regional Contest

• Cryptography

  Grand prize, 2024 Cryptanalysis Contest (암호분석경진대회) (Report)

  3rd Prize, 2013 Korea Cryptography Competition (국가암호공모전)

• Mar. 2017 - Feb. 2022

  Graduate Student Scholarship

  Korea Foundation for Advanced Studies (한국고등교육재단)

• Mar. 2013 - Feb. 2017

  Presidential Science Scholarship (대통령과학장학금)

  Korea Student Aid Foundation (Photo)

Patents

---

As an Inventor

• LOCALITY-PRESERVING MEMORY ARBITER.

  KR Patent 10-2316154 (Application No. 10-2019-0177741), filed December 30, 2019, and issued October 18, 2021.

• METHOD AND SYSTEM FOR PROVIDING CREATING INTERMEDIATE REPRESENTATION.

  KR Patent 10-2467622 (Application No. 10-2022-0072637), filed June 15, 2022, and issued November 11, 2022.

  KR Patent Application 10-2022-0147513, filed November 8, 2022.

  WO Patent Application PCT/KR2022/008505, filed June 15, 2022.

  CN Patent Application CN202280043438, filed June 15, 2022.

• METHOD AND SYSTEM FOR GENERATING INTERMEDIATE REPRESENTATION FOR PROGRAM FOR EXECUTION ON ACCELERATOR.

  KR Patent 10-2457154 (Application No. 10-2022-0072640), filed June 15, 2022, and issued October 17, 2022.

  KR Patent Application 10-2022-0133271, filed October 17, 2022.

  WO Patent Application PCT/KR2022/008508, filed June 15, 2022.

  US Patent Application US18/533,041, filed December 7, 2023.

• METHOD AND SYSTEM FOR MANAGING INTERMEDIATE REPRESENTATION FROM PROGRAM.

  KR Patent 10-2457153 (Application No. 10-2022-0072639), filed June 15, 2022, and issued October 17, 2022.

  KR Patent Application 10-2022-0133270, filed October 17, 2022.

  WO Patent Application PCT/KR2022/008507, filed June 15, 2022.

  US Patent Application US18/542,563, filed December 15, 2023.

• METHOD AND SYSTEM FOR DETERMINING OPTIMIZATION APPLICABILITY ON INTERMEDIATE REPRESENTATION FROM PROGRAM.

  KR Patent 10-2457152 (Application No. 10-2022-0072638), filed June 15, 2022, and issued October 17, 2022.

  KR Patent Application 10-2022-0133269, filed October 17, 2022.

  WO Patent Application PCT/KR2022/008506, filed June 15, 2022.

  US Patent Application US18/542,544, filed December 15, 2023.

• STORAGE SYSTEM FOR QUANTUM CIRCUIT SIMULATION.

  KR Patent Application 10-2022-0043557, filed April 7, 2022.

  WO Patent Application PCT/KR2022/005229, filed April 11, 2022.

• QUANTUM CIRCUIT SIMULATION METHOD AND OPTIMIZATIONS USING STORAGE DEVICES.

  KR Patent Application 10-2022-0043556, filed April 7, 2022.

  WO Patent Application PCT/KR2022/005230, filed April 11, 2022.

• METHOD FOR NEURAL NETWORK QUANTIZATION USING RANDOMIZED SCALE.

  KR Patent Application 10-2021-0012632 filed January 28, 2021.

• METHOD FOR NEURAL NETWORK QUANTIZATION USING BITWIDTH SCHEDULING.

  KR Patent Application 10-2021-0012640, filed January 28, 2021.

• COOLING DEVICE FOR SERVER AND DRIVING METHEOD THEREOF.

  KR Patent Application 10-2020-0160090, filed November 25, 2020.

• COOLING SYSTEM FOR SERVER.

  KR Patent Application 10-2020-0160072, filed November 25, 2020.

• ACCELERATOR, METHOD FOR OPERATING THE SAME AND ACCELERATOR SYSTEM INCLUDING THE SAME.

  KR Patent Application 10-2020-0075655, filed June 22, 2020.

  US Patent Application US17/148,676, filed January 14, 2021.

• APPRATUS AND METHOD OF LARGE SCALE COMPUTING.

  KR Patent Application 10-2021-0185382, filed December 22, 2021.

  US Patent Application US17/901,363, filed September 1, 2022.

  CN Patent Application CN202211055404, filed August 30, 2022.

  EP Patent Application EP22198986, filed September 30, 2022.