Skip to main content

Dr. Chiou Receives NSF CAREER Award

Professor Derek Chiou has been awarded a National Science Foundation CAREER award for his proposal entitled Transforming Computer System Design

The goal of the project is to dramatically improve the process of architecting, implementing and verifying both the hardware, system software and software that make up full computer systems. Traditionally, developing a computer system requires at least three sets of simulators: one that is accurate and flexible for the initial design, one that is accurate and complete as the implementation (RTL) and one that is fast for software development.

The CAREER project is to develop a methodology that will take Dr. Chiou's current research work in a simulator that is simultaneously fast, accurate and flexible and automatically transform it into a complete implementation, thus eliminating two simulators including the most difficult one.  An even more important effect of a single simulator is that each development group will be able to directly affect the others, potentially resulting in more integrated designs.

Dr. Derek Chiou has been awarded a $400K National Science Foundation CAREER award, a prestigious honor which recognizes promising young faculty members. The five-year grant will be used to further develop his proposal entitled Transforming Computer System Design.

Chiou's primary goal is to dramatically improve the expensive and time-consuming process of architecting, implementing and verifying the hardware, system software and application software when building or using computer systems. The CAREER project aims to automatically transform written simulators, using Dr. Chiou's current simulation methodology, into complete implementations.

 

Traditionally, developing a computer system requires at least three separate simulators: one to accurately and flexibly simulate an initial design, a second to accurately implement the system, and a third to quickly develop software applications. If successful, the CAREER grant research would eliminate the need to write two of the simulators, including the most difficult one, the implementation.