Registration: https://events.vtools.ieee.org/m/560442

Date: 4 June 2026

Description: In-person seminar. Social period 5 – 6pm followed by technical session from 6-7pm. Refreshments will be provided (please note dietary restrictions during registration).

Location: Boston Conference Room, Building 6, 1 Analog Way, Wilmington, MA 01887 (All visitors must check in at the first floor of the HUB before crossing over to Building 6)

Speaker: Yizhuo Wu, Delft University of Technology

Topic: Open-Source AI for Analog Correction: From RF Power Amplifiers to Energy-Efficient Silicon

Abstract:

RF PAs dominate the energy budget of modern base stations and Wi-Fi access points, yet their efficiency-oriented designs aggravate nonlinearities. At the same time, next-generation standards demand tighter linearity and modulation accuracy under wide bandwidths, making robust digital predistortion increasingly critical. For practical deployment, two considerations are important: 1) linearity, meeting stringent ACPR and EVM targets even under quantization; 2) energy, keeping inference and update power within tight radio back-end budgets.  This talk will discuss the OpenDPD framework and the role of open-source AI in bridging the gap between algorithm design and silicon deployment for analog/RF systems.

Biography:

Yizhuo Wu received her B.Sc. degree in microelectronics from UESTC, Chengdu, China, in 2021 and her M.Sc. degree in microelectronics at TU Delft in 2023. She is now a Ph.D. student supervised by Dr. Chang Gao in the Lab of Efficient Machine Intelligence. Her research focuses on co–designed software–hardware AI for wireless signal processing, aiming to develop energy-efficient solutions for high-frequency signal processing tasks. She is a recipient of the 2026 IEEE MTT-S Graduate Student Fellowship.