R&D Engineer, Onsemi
Dr. Prateek Sharma received a Ph.D. degree from Vienna University of technology in Semiconductor Device Reliability. He has been working as NVM R&D Engineer at onsemi since 2016. Dr. Sharma’s research has been published in top journals in his field such as IEEE Transactions on Electron Devices, IEEE Electron Device Letters, Microelectronics Reliability and Solid-State Electronics. His work has also been featured at various international conferences, including International Symposium on Power Semiconductor Devices & IC’s, European Solid-State Device Research Conference, International Integrated Reliability Workshop, and International Conference on Simulation of Semiconductor Processes and Devices.
Dr. Sharma is involved in research and development of compute-in-memory technology for low-power edge devices, and IoT (internet of things) devices. Dr. Sharma also worked on AI/ML approaches for optimization of hardware design. Dr. Sharma has patents for an efficient computation memory technique (US20210133550A1) and a software optimization technique using machine learning (ONS03915US).
Dr. Sharma is interested in AI hardware research. He is involved in research and development of Compute-in-memory. This is an innovative memory technology where memory does the storing as well as computing or processing, thus eliminating the need for data to flow between memory and processor as in traditional processors. This technology will make the processing of machine learning information highly efficient (faster and low power). Compute in memory technology has been demonstrated to consume 1,000 times less energy than traditional processors for machine learning inferencing tasks. Applications include object detection in self-driving vehicles, surveillance cameras, low power sensors, edge devices, and IoT (internet of things) devices.