`

Yunho Oh

I am recruiting graduate students (Ph.D. and master). Please contact me if you are interested in research on computer architecture.

I am an assistant professor in the Department of Electronic and Electrical Engineering at SungKyunKwan University (SKKU). My research interests focus on designing next-generation computing systems, including hardware and software architectures for energy-efficient datacenters, processor architectures (CPUs, GPUs, and neural network accelerators), in-storage processing, memory system, and high-performance computing.

I was a postdoctoral researcher in PARSA at EPFL (PI: Professor Babak Falsafi). I completed Ph.D. in the School of Electrical and Electronic Engineering at Yonsei University (Advisor: Professor Won Woo Ro). From August 2016 to August 2017, I worked as a visiting graduate scholar at the University of Southern California (Faculty Sponsor: Professor Murali Annavaram). From 2011 to 2014, I worked as a software engineer at Mobile Communications Buniness, Samsung Electronics.

News

July 2021: I will be serving as a program committee in MLSYS 2022.

June 2021: I will be serving as the poster session chair in ASPLOS 2022.

Mar. 2021: A paper titled "Rebooting Virtual Memory with Midgard" has been accepted to ISCA 2021. Big congrats, Sid!

Mar. 2021: I will be serving as an external review committee in MICRO 2021.

Feb. 2021: I have joined the Department of Electronic and Electrical Engineering at SungKyunKwan University (SKKU) as an assistant professor.

Oct. 2020: Hyunjin has presented a paper titled "Duplo: Lifting Redundant Memory Accesses of Deep Neural Networks for GPU Tensor Cores" in MICRO 2020.

Education

Aug. 2018
Ph.D. in the School of Electrical and Electronic Engineering, Yonsei University, Seoul, Korea

Feb. 2011
M.S. in the School of Electrical and Electronic Engineering, Yonsei University, Seoul, Korea

Feb. 2009
B.S. in the School of Electrical and Electronic Engineering, Yonsei University, Seoul, Korea

Publications

Conferences and Journals

ISCA 2021
Rebooting Virtual Memory with Midgard
Siddharth Gupta, Atri Bhattacharyya, Yunho Oh, Abhishek Bhattacharjee, Babak Falsafi, and Mathias Payer
The 48th ACM/IEEE International Symposium on Computer Architecture, virtual conference, 2021.

MICRO 2020
Duplo: Lifting Redundant Memory Accesses of Deep Neural Networks for GPU Tensor Cores
Hyeonjin Kim, Sungwoo Ahn, Yunho Oh, Bogil Kim, Won Woo Ro, and William J. Song
The 53rd IEEE/ACM International Symposium on Microarchitecture, virtual conference, 2020.

ISCA 2019
Linebacker: Preserving Victim Cache Lines in Idle Register Files of GPUs
Yunho Oh, Gunjae Koo, Murali Annavaram, and Won Woo Ro
The 46th ACM/IEEE International Symposium on Computer Architecture, Phoenix, AZ, USA, 2019.

IEEE TC
Adaptive Cooperation of Prefetching and Warp Scheduling on GPUs
Yunho Oh, Keunsoo Kim, Myung Kuk Yoon, Jong Hyun Park, Yongjun Park, Murali Annavaram, and Won Woo Ro
IEEE Transactions on Computers, vol. 68, no. 4, pp. 609-616, April. 2019.

MICRO 2018
FineReg: Augmenting GPU Throughput via Fine-Grained Register File Management
Yunho Oh, Myung Kuk Yoon, William J. Song, and Won Woo Ro
The 51st ACM/IEEE International Symposium on Microarchitecture, Fukuoka, Japan, 2018

IEEE TC
WASP: Selective Data Prefetching with Monitoring Runtime Warp Progress on GPUs
Yunho Oh, Myung Kuk Yoon, Jong Hyun Park, Yongjun Park, and Won Woo Ro
IEEE Transactions on Computers, vol. 67, no. 9, pp. 1366-1373, Sept. 2018.

IEEE TPDS
Dynamic Resizing on Active Warps Scheduler to Hide Operation Stalls on GPUs
Myung Kuk Yoon, Yunho Oh, Sangpil Lee, Seung Hun Kim, Deokho Kim, and Won Woo Ro
IEEE Transactions on Parallel and Distributed Systems, vol. 28, no. 11, pp. 3142-3156, Nov. 2017.

ISCA 2017
Access Pattern-Aware Cache Management for Improving Data Utilization in GPU
Gunjae Koo, Yunho Oh, Won Woo Ro, and Murali Annavaram
The 44th ACM/IEEE International Symposium on Computer Architecture, Toronto, ON, Canada, 2017.

ISCA 2016
APRES: Improving Cache Efficiency by Exploiting Load Characteristics on GPUs
Yunho Oh, Keunsoo Kim, Myung Kuk Yoon, Jong Hyun Park, Yongjun Park, Won Woo Ro, and Murali Annavaram
The 43rd ACM/IEEE International Symposium on Computer Architecture, Seoul, Korea, 2016.

ISPASS 2015
DRAW: Investigating Benefits of Adaptive Fetch Group Size on GPU
Myung Kuk Yoon, Yunho Oh, Sangpil Lee, Seung Hun Kim, Deokho Kim, and Won Woo Ro
The 2015 IEEE International Symposium on Performance Analysis of Systems and Software, Philadelphia, PA, USA, 2015.

ITC-CSCC 2015
Improving Pipeline Utilization with Two-Level Instruction Issue on GPUs
Yunho Oh, Jong Hyun Park, and Won Woo Ro
The 30th International Technical Conference on Circuits/Systems, Computers and Communications, Seoul, Korea, 2015.

KIISE
Introduction to Researches on Performance Bottlenecks of Many-Core GPU Architectures
Yunho Oh, Myung Kuk Yoon, Jong Hyun Park, and Won Woo Ro
Communications of KIISE, Vol. 32 No. 5, May, 2014.

IJPP
GPU-Friendly Parallel Genome Matching with Tiled Access and Reduced State Transition Table
Yunho Oh, Doohwan Oh, and Won W. Ro
International Journal of Parallel Programming, Vol. 41, No. 4, pp. 526-551, August, 2013.

ICEIC 2010
Multi-Threaded Filtered BackProjection Algorithm on Multi-Core Processors
Yun H. Oh, and Won W. Ro
The 10th International Conference on Electronics, Information, and Communication, Cebu, Philippines, 2010.

ISMRM 2010
Accelerated Reconstruction Using Parallel Computing for Spiral Spectroscopic Imaging
Dong H. Kim, Yun H. Oh, Yun H. Nam, M. Gu, and Won W. Ro
International Society for Magnetic Resonance in Medicine Annual Meeting, Stockholm, Sweden, 2010.

ELEX
Hardware Implementation of a Tessellation Accelerator for the OpenVG Standard
Seung Hun Kim, Yunho Oh, Karam Park, and Won W. Ro
IEICE Electronics Express, Vol. 7, No. 6, pp 440-446, March, 2010.

Workshops

SPMA
AstriFlash: An Online Flash-Based Memory Hierarchy
Siddharth Gupta, Yunho Oh, Lei Yan, Mark Sutherland, Abhishek Bhattacharjee, Babak Falsafi, and Peter Hsu
The 10th Workshop on Systems for Post-Moore Architectures co-located with Eurosys 2020.

HENND
Accelerating Neural Network with Selective Thread-Level Parallelism Regulation and Cache Bypassing on GPUs
Kwanghee Chang, Yunho Oh, Myung Kuk Yoon, and Won Woo Ro
International Workshop on Highly Efficient Neural Networks Design in conjunction with ESWEEK 2017.

Patent

Central processing unit, GPU simulation method thereof, and computing system including the same
Won Woo Ro, Karam Park, Yunho Oh, Sangpil Lee, and Minwoo Kim
US Patent 9,378,533.

Experience

Professional Experience

Feb. 2021 - Current  Assistant Professor, SungKyunKwan University  

Mar. 2019 - Feb. 2021  Postdoctoral Researcher, École Polytechnique Fédérale de Lausanne  
Principal Investigator: Prof. Babak Falsafi
- New memory system architecture for cost-efficient datacenters
- Microarchitecture-specific server workload behavior analysis
- Scalable neural network accelerator design

Sep. 2018 - Jan. 2019.  Postdoctoral Researcher, Yonsei University  
- New register file architecture for GPUs with new memory technology
- Energy-efficient neural network processing on GPUs
- Entropy analysis on GPU register compression

Mar. 2014 - Aug. 2018.  Graduate Research Assistant (Ph.D.), Yonsei University  
- Memory locality-aware warp scheduling technique on GPUs
- Efficient prefetching techniques on GPUs
- Efficient register file management on GPUs
- Energy-efficient memory system for GPUs
- Energy-efficient multi-core processor architecture

Aug. 2016 - Aug. 2017.   Visiting Graduate Scholar, University of Southern California, LA, CA  
Faculty Sponsor: Prof. Murali Annavaram
- Efficient cache management technique on GPUs
- In-SSD database index management system

Feb. 2011 - Jan. 2014.   Software Engineer, Mobile Communications Business, Samsung Electronics  
- Developed mobile applications for Android smartphones
- Developed a cloud service framework for wearable devices

Mar. 2009 - Feb. 2011.   Graduate Research Assistant (M.S.), Yonsei University  
- Parallel genome sequence matching algorithm for GPUs
- High-performance medical image construction algorithms
- OpenVG accelerator prototype design

Teaching Experience

Fall, 2021.   Embedded System Designs
Sungkyunkwan University, Suwon, South Korea

Spring, Summer, and Fall, 2021.   Logic Design Laboratory
Sungkyunkwan University, Suwon, South Korea

Fall, 2020.   Topics on ML Systems (CS723)
EPFL, Lausanne, VD, Switzerland
(I ran a seminar course with student presentations followed by an in-class discussion with Prof. Babak Falsafi, Prof. Anne-Marie Kermarrec, and Prof. Martin Jaggi)

Spring, 2020.   Topics on Datacenter Design (CS728)
EPFL, Lausanne, VD, Switzerland
(I ran a seminar course with student presentations followed by an in-class discussion with Prof. Babak Falsafi and Prof. Anne-Marie Kermarrec)

Fall, 2019.   Semester Project: I co-supervised an undergraduate student with Dr. Mirjana Stojilovic
EPFL, Lausanne, VD, Switzerland
(GPU acceleration of electromagnetic time-reversal algorithm)

Fall, 2019.   Introduction to Multiprocessor Architecture (CS307)
EPFL, Lausanne, VD, Switzerland
(I gave lectures regarding cache-coherence protocols, multi-threaded processors, and GPUs)

Dec. 2015.   GPGPU Programming Practice
Samsung Advanced Technology Training Institute (SAIT)

Spring, 2010.   Teaching Assistant, EEE3530: Computer Architecture
Yonsei University

Fall, 2010.   Teaching Assistant, EEE2111: Electrical and Electronic Engineering Experiments
Yonsei University

Talks

Aug. 2021.   Data-Centric Computing, Brick by Brick
2021 SIGARCH Korea Workshop (Virtual)

Aug. 2021.   Towards Energy-Efficient DNN Acceleration Systems
2021 Summer School, Korean Society of Artificial Intelligent in Medicine (Virtual)

June. 2021.   Challenges in Next-Generation Datacenters: Accelerators and Memory Systems
IT21 Conference (Virtual)

June. 2020.   Designing Workload-Aware GPU Architectures 
HiSilicon (Remote talk)

Sep. 2019.   Resolving Critical GPU Performance Bottlenecks 
Hongik University, Seoul, Korea

Nov. 2018.   Rethinking GPU Execution Model 
EPFL, Lausanne, VD, Switzerland

Oct. 2018.   Towards Extreme Throughput 
EPFL, Lausanne, VD, Switzerland

Dec. 2018.   FineReg: Augmenting GPU Throughput via Fine-Grained Register File Management 
Invited Talk in Software Congress 2018, PyeongChang, Korea

Service

Organization Committee
ACM International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), 2022

Program Committee
Conference on Systems and Machine Learning (MLSys), 2022
Workshop on the Computing Continuum (in HiPEAC), 2020

External Review Committee
ACM/IEEE International Symposium on Microarchitecture (MICRO), 2021
IEEE/ACM International Symposium on Microarchitecture (MICRO), 2020
ACM/IEEE International Symposium on Computer Architecture (ISCA), 2019

Reviewer
IEEE Transactions on Emerging Topics in Computing, 2021
IEEE Computer Architecture Letters, 2021
Future Generation Computer Systems, 2020
IEEE Computer Architecture Letters, 2019
IEEE Access, 2019

Honors and Awards

Nov. 2017.   Excellent Graduate Researcher Scholarship 
Yonsei University

Aug. 2016 - Aug. 2017.   BK21 Long-Term Overseas Research Fellowship 
Yonsei University

Mar. 2009 - Feb. 2011.   Mobile Communication Graduate Fellowship 
Samsung Electronics

Feb. 2008.   Excellent Undergraduate Awarded Honors and Scholarship 
Yonsei University

Contact

E-mail   [firstname].[lastname] at skku dot edu

LinkedIn   https://www.linkedin.com/in/yunho-oh-comarch/