1. Device-Centric Server Architecture
In this project, we design and prototype innovative server architectures to satisfy the need of emerging server applications (e.g., data analytics, machine learning, virtual machine). To achieve the applications’ computation, memory, and storage needs, we provide full-system, scale-up, and scale-out server architecture solutions by actively exploiting FPGA-assisted device orchestration and acceleration methods.
2. Cryogenic/Superconducting/Quantum Computer System
In this project, we design and build various cryogenic computer devices and systems. Cryogenic computers run at extremely low temperatures (e.g., 4K, 77K) to dramatically reduce the wire latency and leakage current. To take the best advantages of cryogenic computing, we develop a cryogenic computer modeling tool and provide our cryogenic processor, cache, memory, and server designs.
3. AI/Deep Learning Acceleration System
In this project, we analyze the need of the current industry-hot AI applications (e.g., CNN, RNN, MANN) and provide various architecture and system solutions to efficiently accelerate them. Our current interests lie in developing highly-scalable and flexible AI acceleration systems by actively exploiting heterogeneous hardware solutions together (e.g., FPGA, ASIC, GPU, SSD).
4. Neuromorphic Computing and Brain-Computer Interface System
In this project, we pursue two complementary directions to better understand and interact with the brain. We build a fast and scalable neuromorphic computing system to faithfully simulate human-scale neural circuits and uncover how spike-based networks enable cognition. In parallel, we develop scalable and efficient brain–computer interface (BCI) systems that can reliably interact with the brain for practical applications.
5. Computer System Modeling and Analysis
In this project, we develop and release various computer system modeling and analysis methodologies. Our methodologies include efficient computer architecture simulation methods (e.g., performance, power consumption) and performance analysis methods to understand various commodity computer systems whose internal mechanisms are undisclosed (e.g., processor, storage, datacenter).