A paper published by a team of University of Utah electrical and computer engineering researchers led by professor Priyank Kalla recently received the honorable mention award at the International Test Conference in India. The paper, titled “Transfer-Matrix Abstractions to Analyze the Effect of Manufacturing Variations in Silicon Photonic Circuits”, was a collaborative effort between Kalla’s research group and professor Steve Blair, primarily authored by Ph.D. student Pratishtha Agnihotri.

“Photonic chips are very sensitive to variations in the manufacturing process. If there are small fabrication errors, the devices may not work the way we want them to, and it could cause errors in computation,” Kalla said. “When you integrate a large number of optical devices, it becomes difficult to analyze which device is not working properly. The analysis requires models that are computationally very expensive and essentially infeasible.”

The paper investigates mathematical abstractions with the goal of developing models that can automate quality testing for larger circuits. Ultimately, the group hopes to streamline the process of quality-checking and troubleshooting large integrated optoelectronic systems.

“We want to develop CAD tools which will be able to analyze production quality of optical circuits,” says Kalla. “These tools will allow us to determine the effect of manufacturing variations on circuit performance, and further in the future we should be able to use this technology to speed-up and lower the cost of testing of a circuit before it is shipped to the market.”

Kalla and Blair have been working on this project for several years now, with the initial design phase beginning back in 2013. Agnihotri joined Kalla’s lab as a Ph.D. student in January of 2021 and is currently the sole individual in the research group working in this domain.

“This was my first paper, so it was really exciting to receive an award for the first initiative I worked on,” says Agnihotri. “Photonics Design Automation is a very new domain, so we want to develop these tools quickly.”

Agnihotri recently presented the awarded paper at ITC USA last month. To read the team’s full award-winning work, click here.