Ranggi Hwang
RESEARCH INTERESTS
Computer architecture and system
Machine learning accelerator
Hardware-software co-design for machine learning
EDUCATION
Ph.D. in Electrical Engineering, March 2021 - Present
M.S. in Electrical Engineering, March 2019 - February 2021
Dissertation : Hybrid CPU-FPGA Accelerator for Deep Learning based Personalized Recommendations
Advisor: Prof. Minsoo Rhu
B.S. in Electrical Engineering, March 2013 - February 2019
RESEARCH EXPERIENCE
Graduate research assistant @ Vertically Integrated Architecture Research Group, KAIST
March 2019 - Present
Undergraduate research assistant @ Mobile Multimedia System Laboratory, Seoul National University
July 2018 - September 2018
Assisted with research on low computational complexity algorithm of Spiking Neural Network
Researched hardware-friendly Spike-Timing-Dependent Plasticity, neuron and synapse modeling
Undergraduate research assistant @ Bio-Application System and Integrated Circuit Laboratory, Korea University
March 2018 - June 2018
Researched in-vitro simulation system of implantable neural interface IC
WORK EXPERIENCE
Research Intern @ Microsoft Research Asia
October 2022 - April 2023
Research on the GPU acceleration of LLM inference
(Hwang et al., "Pre-gated MoE: An Algorithm-System Co-Design for Fast and Scalable Mixture-of-Expert Inference," ISCA-2024)Advisor: Dr. Shijie Cao
Research Engineer Intern @ Samsung Electronics
January 2018 - February 2018
Researched connectivity and power management for visual display circuit
Participated prototype development for UHD TVs and outdoor display products
PUBLICATION
Ranggi Hwang*, Jianyu Wei*, Shijie Cao, Changho Hwang, Xiaohu Tang, Ting Cao, and Mao Yang,
"Pre-gated MoE: An Algorithm-System Co-Design for Fast and Scalable Mixture-of-Expert Inference,"
The 51st IEEE/ACM International Symposium on Computer Architecture (ISCA-51), Buenos Aires, Argentina, June 2024Juntaek Lim, Youngeun Kwon, Ranggi Hwang, Kiwan Maeng, Edward Suh, and Minsoo Rhu,
"LazyDP: Co-Designing Algorithm-Software for Scalable Training of Differentially Private Recommendation Models,"
The 29th ACM International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS-29), San Diego, CA, USA, Mar. 2024Acceptance Rate: 20% (193 among 921)
[Paper] [Presentation] [Code]
Ranggi Hwang*, Minhoo Kang*, Jiwon Lee, Dongyun Kam, Youngjoo Lee, and Minsoo Rhu,
"GROW: A Row-Stationary Sparse-Dense GEMM Accelerator for Memory-Efficient Graph Convolutional Neural Networks,"
The 29th IEEE International Symposium on High-Performance Computer Architecture (HPCA-29), Montreal, QC, Canada, Feb. 2023* Co-first authors
Acceptance Rate: 25% (91 among 364)
[Paper]
Seonho Lee, Ranggi Hwang, Jongse Park, and Minsoo Rhu,
"HAMMER: Hardware-friendly Approximate Computing for Self-attention with Mean-redistribution and Linearization,"
IEEE Computer Architecture Letters, Jan. 2023[Paper]
Beomsik Park*, Ranggi Hwang*, Dongho Yoon, Yoonhyuk Choi, and Minsoo Rhu,
"DiVa: An Accelerator for Differentially Private Machine Learning,"
The 55th IEEE/ACM International Symposium on Microarchitecture (MICRO-55), Chicago, IL, USA, Oct. 2022Ranggi Hwang, Taehun Kim, Youngeun Kwon, and Minsoo Rhu,
"Centaur: A Chiplet-based, Hybrid Sparse-Dense Accelerator for Personalized Recommendations,"
The 47th IEEE/ACM International Symposium on Computer Architecture (ISCA-47), Valencia, Spain, June 2020Acceptance Rate: 18% (77 among 421)
[Paper] [Presentation] [Article]
HONOR AND AWARD
Student travel grant, The 51st IEEE/ACM International Symposium on Computer Architecture (ISCA-51), 2024
Lee Moon-Chi Scholarship, College of Engineering, Korea university, 2018
Semester High Honors, College of Engineering, Korea University, 2017, 2018
Office of Student Affairs Special Scholarship, Korea University, 2017, 2018
Veritas Scholarship, Korea University, 2017
Volunteer Scholarship, Korea University, 2014
TECHNICAL SKILLS
Programming Languages : Verilog HDL, SystemVerilog, Python, C, C++, CUDA, ARM Assembly (ADS), MIPS, R
Machine Learning Frameworks : PyTorch, TensorFlow, Caffe
Others : Spice, Quartus, Multisim, MATLAB
ACADEMIC SERVICE
TEACHING EXPERIENCE
Teaching Assistant (Head TA), EE485: Software Development Environment and Tools Practice, Fall 2023 @KAIST
Teaching Assistant, EE209: Programming Structure for Electrical Engineering, Spring 2022 @KAIST
Teaching Assistant, EE312: Introduction to Computer Architecture, Fall 2021 @KAIST
Teaching Assistant, EE511: Computer Architecture, Spring 2021 @KAIST
LANGUAGE PROFICIENCY
Native in Korean, Fluent in English