PGI-14 hosted Seminar - Dr. Thomas Van Vaerenbergh (Hewlett Packard Laboratories, HPE)
Dear colleagues,
We cordially invite you to the PGI-14 hosted Seminar on 17th August 2022, 10:00 AM.
Please feel free to forward to those in your team who might be interested.
Additionally, Dr Van Vaerenbergh is pleased to meet with prospective collaborators or interested parties during his time at Jülich/Aachen. Any individuals or groups interested to meet on August 17 can send a brief note to: j.strachan@fz-juelich.de or p.wagner@fz-juelich.de. It may also be possible to arrange ad-hoc meetings following Dr Van Vaerenbergh’s seminar.
The talk will take place as a hybrid event with a reduced number of seats in the PGI lecture hall (Building 04.8, 2nd floor, Room 365) and a video conference in parallel. To join the talk online please use the dial-in option for the video conference:
https://bluejeans.com/932703754/6183
Dr. Thomas Van Vaerenbergh (Hewlett Packard Labs, HPE, Brussels)
will give a talk on:
Silicon photonics for energy-efficient optical interconnects and optical computing
Abstract: To keep up with the latency, energy-efficiency and bandwidth needs of future data center and high-performance computing workloads in a cost-efficient way, Hewlett Packard Labs is currently developing silicon photonic interconnects that can hit the required specifications. In this talk, I will give an overview of HPE’s recent developments on heterogeneous III/V-on-Si integration lead to sub-pJ/bit energy-efficiency for dense wavelength division multiplexed (DWDM) transceivers [1].
In addition, I will illustrate how these and related advances, such as the development of non-volatile tuning based on the memristive effect [2], make the hybrid III/V-on-Si material stack a promising platform for analog photonic accelerators targeting machine learning [3] and combinatorial optimization workloads [4]. I will conclude the talk by explaining how machine learning techniques can be utilized as part of inverse design routines for photonic devices.

Bio:
Thomas Van Vaerenbergh received his Master's degree in Applied Physics and his Ph.D. in Photonics from Ghent University, Ghent, Belgium, in 2010 and 2014, respectively. He was awarded the scientific prize Alcatel-Lucent Bell/FWO for his Ph.D. thesis on all-optical spiking neurons in silicon photonics. In 2014, he joined the Palo Alto-based division of the Large-Scale Integrated Photonics team in Hewlett Packard Labs, part of Hewlett Packard Enterprise (HPE).
Since 2019, he has been working HPE Belgium’s office, with the aim to expand HPE’s research activities related to photonics and AI in the EMEA region. His main research interests include optical computing, analog photonic and electronic accelerators for combinatorial optimization, the modeling and design of passive silicon photonic devices, such as microring resonators and grating couplers, and the usage of state-of-the-art machine learning techniques to facilitate the design of photonic devices, circuits and devices.
Bibliography:
[1] D. Liang et al., "An Energy-Efficient and Bandwidth-Scalable DWDM Heterogeneous Silicon Photonics Integration Platform," in IEEE Journal of Selected Topics in Quantum Electronics, vol. 28, no. 6, pp. 1-19, Nov.-Dec. 2022, Art no. 6100819, doi: 10.1109/JSTQE.2022.3181939.
[2] B. Tossoun, X. Sheng, J. P. Strachan and D. Liang, "The Memristor Laser," 2020 IEEE International Electron Devices Meeting (IEDM), 2020, pp. 7.6.1-7.6.4, doi: 10.1109/IEDM13553.2020.9371989.
[3] Xian Xiao, Mehmet Berkay On, Thomas Van Vaerenbergh, Di Liang, Raymond G. Beausoleil, and S. J. Ben Yoo , "Large-scale and energy-efficient tensorized optical neural networks on III–V-on-silicon MOSCAP platform", APL Photonics 6, 126107 (2021) https://doi.org/10.1063/5.0070913
[4] N. Tezak et al., "Integrated Coherent Ising Machines Based on Self-Phase Modulation in Microring Resonators," in IEEE Journal of Selected Topics in Quantum Electronics, vol. 26, no. 1, pp. 1-15, Jan.-Feb. 2020, Art no. 5900115, doi: 10.1109/JSTQE.2019.2929184.