Room 1037 (Emerging Technology Building- ETB)
Zhiyu (Albert) Zeng
(Samsung Austin R&D Center)
Due to recent aggressive process scaling into the nanometer regime, power delivery network design faces many challenges such as large sheer network complexity and the employment of various cutting-edge low-power techniques. All these challenges set more stringent and specific requirements to the EDA tools for power delivery networks. For example, on the analysis perspective, simulation efficiency for large grids must be improved. Also, good power delivery network designs need to have the optimal parameters assigned to various system components such as decaps, voltage regulators and converters. This presentation proposes new methodologies to address these challenging problems. At first, a novel parallel partitioning-based approach for static analysis is presented. This approach provides a flexible network partitioning scheme and a block-based iterative error-reduction flow. Later, a systematic design analysis is conducted on power delivery networks that incorporate buck converters and low-dropout voltage regulators. The electrical interactions between active regulators/converters and passive networks, and their influences on key system design specifications are analyzed. With the derived design insights, the system-level co-design of a complete power delivery network is facilitated by an automatic optimization flow.
Bio: Zhiyu (Albert) Zeng received the B.S. degree in electronic and information engineering from Zhejiang University, China, in 2006 and the Ph.D. degree in computer engineering from Texas A&M University in 2011. He was the recipient of the 3rd Prize in 2011 TAU Power Grid Simulation Contest. He is now a CAD engineer in Samsung Austin R&D Center. His research interests include parallel simulation and optimization of power grids and statistical SRAM analysis.