김장우 연구실

Design and prototype innovative server architectures for emerging applications (data analytics, machine learning, virtual machine) using FPGA-assisted device orchestration and acceleration methods.

Design and build cryogenic computer devices and systems operating at extremely low temperatures (4K, 77K) to reduce wire latency and leakage current, with cryogenic processor, cache, memory designs.

Analyze and provide architecture and system solutions to efficiently accelerate AI applications (CNN, RNN, MANN) using heterogeneous hardware solutions (FPGA, ASIC, GPU, SSD).

Design and build fast, efficient, scalable neuromorphic computer system to simulate brain using spike-based networks for brain-inspired computing.

Develop computer system modeling and analysis methodologies including efficient architecture simulation (performance, power consumption) and performance analysis for commodity systems.