6/5/2013 Tom Moone, Computer Science
Written by Tom Moone, Computer Science
CS Professor and Department Head Rob Rutenbar, faculty in the Coordinated Science Laboratory, was selected as one of the recipients of the 2013 Donald O. Pederson Best Paper Award. Recognizing an article published in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, this annual award is given to a significant research article published in the journal during the preceding two calendar years. Rutenbar was a co-author of “Virtual Probe: A Statistical Framework for Low-Cost Silicon Characterization of Nanoscale Integrated Circuits,” which appeared in the December 2011 issue of the journal.
“IEEE Transactions on CAD is the most elite journal in this area,” said Rutenbar, an Abel Bliss Professor in Engineering at Illinois. “It’s kind of the biggest best paper award in the CAD business.”
The research presented in the article began when Rutenbar read about work on compressive sensing, which was a breakthrough in mathematics. “It gave the ability to re-create a scene from a remarkably small number of observations based on a few assumptions about the mathematical regularity of what you are looking at,” said Rutenbar.
Rutenbar discussed this work with his colleague Xin Li, one of the co-authors of the paper, and how they might use compressive sensing techniques in chip design—in particular for measurements on a wafer’s surface. Such measurements on the surface, called probes, are an important part of the manufacturing process. Even though every chip on a wafer is supposed to be exactly the same, the physics of manufacturing means that there is variation over the surface of the wafer. “You need to be able to measure stuff to be able to predict what is and is not working on the wafer,” said Rutenbar.
However, probing the surface of the wafer can be a time-consuming, expensive process. Rutenbar and his colleagues came up with what they called a “virtual probe.” Rutenbar explained, “You put a small number of physical probes down, and then using reconstructive mathematics, you can predict results at other locations that you did not physically probe.”
The process could have implications in industry. Several of Rutenbar’s co-authors were from IBM, which provided actual industrial data for use in the research. “The people from IBM were excited by the opportunities presented by the technologies,” said Rutenbar. “We’re hoping there’s going to be some uptake.”
Rutenbar’s more recent research interests have focused on developing custom computer architectures for difficult problems and on machine learning and statistical inference. Though he is not working directly on developing tools like the one which led to the Pederson Award, “the math that describes the things I’m designing now is basically what we won the best paper award for,” he said.
Rutenbar is a Fellow of IEEE and ACM. He received the Semiconductor Research Corporation Aristotle Award in 2001 and the IEEE Circuits and Systems Industrial Pioneer Award in 2007. Rutenbar’s first startup company, Neolinear, was acquired by Cadence (NASDAQ: CDNS) in 2004. At his second startup, Voci Technologies, he currently serves as Chairman and Chief Scientist.