ARCS

RESEARCH

Our research has focused on system optimizations for interesting applications in modern eras, which include deep learning frameworks and big data analysis platforms. Our current interests include:

• Compiler Optimization for GPU/FPGA Computing
• Platform Optimization for Big Data/Machine Learning Frameworks
• System Software Design for Non-Volatile Memories
• Static/Dynamic Program Analysis for Performance and Security

Current Grants

• Loop analysis based optimizations for deep learning compilers, Samsung Advanced Institute of Technology (SAIT), 2023.03-2024.02
  • Investigate loop analysis for general fusion/fission methods integrated with schedulers for multiple devices.
• Analysis of applications for zns-ssd storage systems and application, Samsung Electronics, 2022.06-2024.05
  • Investigate application characteristics for specialized-SDDs and explore design space.
• Improving FPGA Programmability for Big Data Analytics, National Research Foundation (NRF), 2021.03-2025.02
  • Investigate techniques for accelerator programmability on big data analytics frameworks.
• Research on Edge-Native Operating Systems for Edge Micro-Data-Centers, Institute of Information & Communications Technology Planning & Evaluation (IITP), 2021.04-2024.12
  • Develop abstraction techniques in compilers for HW-accelerators in EMDC operating systems.

[ Past Grants ]