Faculty

Michael Huang, Ph.D

Assistant Professor
Ph.D., University of Illinois at Urbana-Champaign, 2002

Key Words:
Computer Architecture, Processor Microarchitecture, Energy-Efficient System, Program Behavior Analysis

Research Interests:

Energy-Efficient System Design- Power consumption and heat dissipation have quickly become the central concern of designing modern high-performance processors. In the past, we have investigated an array of issues in designing energy-efficient circuit, logic, microarchitecture, and computer systems. We are looking at energy-efficient solutions for maintaining high system integrity for processors and systems built in deep submicron technologies.

Self-Tuning Systems ­ As general-purpose computers are used in ever wider application domains, improving system quality (performance, energy efficiency, and cost effectiveness) becomes increasingly application-dependent. Moreover, as computer applications become more and more powerful and sophisticated, their behavior becomes complex and dynamic during execution. Our goal is to develop systematic mechanisms to capture program behavior changes and predict future behavior in order to allow the computing system to adapt to the changing program behavior. In other words, the system performs self-tuning.

Resource-Effective Microprocessor Design - Current out-of-order microprocessors routinely employ large structures for nearly all components such as issue queue, load-store queue. Larger structures increase the processor's ability to exploit ILP (instruction-level parallelism). However, accessing large structures takes longer and costs more energy. We look at exploiting high-level knowledge of program behavior to reduce hardware resource demand.

Research Labs/Centers:

•	Advanced Computer Architecture Laboratory
•	Computer Circuits and Systems Group