For pioneering contributions to low-power computer architecture and its interaction with technology.Scroll Up
Trevor Mudge of the University of Michigan received the ACM - IEEE CS Eckert-Mauchly Award in 2014 for contributions to low-power computer architecture for high-performance microprocessors. Mudge’s inventive approaches have led to new technologies that reduce energy consumption of microprocessors while maintaining acceptable performance in an era of exponential growth in embedded processors and system-on-chip designs. His contributions greatly influenced both the research literature and the actual products made possible by his research.
During his years as a research professor, Mudge recognized that limiting power consumption presented a critical computing issue. He concluded that reducing power consumption would require adding architectural improvements to process and circuit improvements, and raised the priority of this constraint early in the design stage. He also understood that, as the speed of microprocessors increased along with density, their leakage power consumption also increased. This realization led him to identify on-chip caches as one of the main candidates for leakage reduction since they contain a significant fraction of the processor’s transistors. The resulting technique for reducing leakage power was the exploitation of “drowsy caches,” which led to putting the cold cache lines into a low-power mode.
With his team from the University of Michigan, Mudge proposed Razor, a circuit technique that allows robust operation at very low voltages in processor pipelines based on dynamic detection and correction of circuit timing errors. More recently, he has explored near-threshold computing, a design space where the supply voltage is approximately equal to the threshold voltage of the transistors in a microprocessor. The approach is applicable to a broad range of power-constrained computing segments from sensors to higher-performance servers.
A graduate of the University of Reading, England with a B.S. degree in Cybernetics, Mudge earned M.S. and Ph.D. degrees in Computer Science from the University of Illinois. He holds the Bredt Family Chair of Engineering at the University of Michigan.