Sunpyo Hong







About me
I joined Intel's Technology Pathfinding and Innovations (TPI) group as a software engineer in Jan 2013. I work on research and advanced development of future software and architectural tools, by focusing on parallel performance and energy efficiency, GPGPU, static/dynamic instruction analysis, heterogeneous scheduling, and many-core architectures

Research
My reseach group in school was HPArch, led by Prof. Hyesoon Kim. I received the Bachelor's Degree in Electrical and Computer Engineering from Georgia Tech in 2002, Master's degree in 2008, and Ph.D. in 2012. I have published in top computer-architecture conferences including ISCA and MICRO, concentrating on GPU performance and energy consumption. I further participated on the research of energy-efficient heterogeneous scheduling and thermal analysis.
Education
Georgia Institute of Technology, Atlanta, GA
Ph.D. in Electrical and Computer Engineering, 2009 - 2012
M.S. in Electrical and Computer Engineering, 2007 - 2008
B.S. in Electrical and Computer Engineering, 1998 - 2002
Publications
  • Melanie Kambadur, Sunpyo Hong, Juan Cabral, Harish Patil, Chi-Keung Luk, Sohaib Sajid, Martha A. Kim
    "Fast Computational GPU Design with GT-Pin". IISWC 2015

  • Sunpyo Hong and Hyesoon Kim
    "A Memory-level and Thread-level Parallelism Aware GPU Architecture Performance Analytical Model",
    In Proc. of the 36th Int'l Symp. on Computer Architecture [ISCA-36], Austin, Texas, USA, June, 2009.
    [PDF] [PPT]

  • Sunpyo Hong and Hyesoon Kim
    "An Integrated GPU Power and Performance Model",
    In Proc. of the 37th Int'l Symp. on Computer Architecture [ISCA-37], Saint-Malo, France, June, 2010
    [PDF] [PPT]

  • Chi-Keung Luk, Sunpyo Hong and Hyesoon Kim
    "Qilin: Exploiting Parallelism on Heterogeneous Multiprocessors with Adaptive Mapping",
    In Proc. of the 42nd Int'l Symp. on Microarchitecture [MICRO-42], New York, USA, Dec, 2010
    [PDF]

  • Sunpyo Hong, Aniruddha Dasgupta, Hyesoon Kim and Jinil Park
    "A New Fine-Scaled Thermal-Analysis System For GPU Architectures using Thermocouples",
    Submitted to IEEE TCAD

  • Sunpyo Hong, Joo Hwan Lee and Hyesoon Kim
    "An OpenCL Throughput Model for Energy-Efficient and High-Performance Work Partitioning",
    Submitted to ISCA

  • Sunpyo Hong and Hyesoon Kim
    "Parallelization of Mutual-Information based Registration in the ITK Toolkit using CUDA and TBB",
    Technical Report, Georgia Institute of Technology, Dec, 2008
    [PDF]

Google Scholar
Thesis
Modeling Performance and Power for Energy-Efficient Computing
CV
Contact
Email: sunpyo.hong at sphong dot net



Last modified: Apr 2016.