The inconsistent performance of photonic and electronic integrated circuits is the focus of Early CAREER award recipient Zheng Zhang, an assistant professor at the University of California, Santa Barbara. Because the nanoscale fabrication process is difficult to control, some semiconductor chips work well, some underperform, and others don’t work at all. Through uncertainty-aware design automation — where the approach is to expect the unexpected and adjust ahead of time — Zhang and his students hope to make semiconductor manufacturing more efficient and the products more consistent. Assistant professor Zheng Zhang. Courtesy of UCSB. “The short-term goal is to develop efficient algorithms to model, verify, and optimize the uncertain performance of an electronic or photonic chip before it’s being fabricated,” Zhang said. “This will significantly improve the product yield of a semiconductor company by increasing the percentage of successful chips in mass production, which will benefit consumers.” Long term, Zhang is interested in extending his methodology to other domains such as robotics and self-driving automobiles. The National Science Foundation award includes up to $500,000 in funding for Zhang to pursue research into how to quantify uncertainty.