Would you like to perform cutting-edge research at the
intersection between IC hardware and artificial intelligence? Do you want to explore a
new class of electronic-photonic components that mimics how biological neurons behave and helps reduce energy consumption and increase performance for AI hardware?
InformationBackgroundRapid advances in artificial intelligence technologies have led to powerful models and algorithms that have revolutionized many applications across all fields of science and technology. Deep learning performed within artificial neural networks has yielded new ways to process data, leading to sophisticated systems with impressive functionality and benefits. However, conventional computing hardware is reaching its limits in terms of energy efficiency and speed. A new approach to computing hardware is needed.
Novel neuromorphic chips working with brain-inspired spiking neural networks have gained attention as they promise highly efficient ways to process data. Important research effort has been dedicated to develop such neuromorphic systems in electronic and photonic hardware. We offer a research position in this fast-paced research field, embedded within a world-class research group.
The integrated photonics fieldSimilar to electronic ICs, photonic ICs are revolutionizing areas such as healthcare, communication and sensing, and have the potential to be disruptive to the whole society. Novel PIC components will have a big impact on the fields of sensing and processing. A diversity of PIC-based sensors have been proposed in the last years, such as environmental sensors (e.g. gas sensing), medical sensors (e.g. optical coherence tomography), fiber Bragg grating sensors for structural monitoring (temperature or strain measurements), light detection and ranging (LiDAR) and others. Also, novel circuit architectures are being proposed for performing computation and the processing of real-time sensory data at the edge.
Recently, a new field of neuromorphic photonics is emerging, which aims to build artificial opto-electronic neurons that mimic the brain for processing information based on synaptic processes. Taking advantage of their threshold-based characteristics, neuromorphic photonic devices can also be used for spike-based processing and event-based sensing to allow recognition of patterns in an ultra-fast and energy-efficient manner.
We have made initial steps towards implementing opto-electronic neurons in InP integrated photonics. We demonstrated a first generation of resonant tunneling diodes (RTDs), which display well-defined electrical excitable characteristics via tunneling effects through a single quantum-well. Full opto-electronic neuromorphic behavior has been demonstrated by coupling the RTDs with external lasers and photodetecting devices in cooperation with partners. See
M. Hejda, ACS Photonics 2024, 11(10) [1].
We are now pursuing the monolithic integration of RTDs with integrated photonic devices to form electronic-photonic neurons on chip, thereby setting the grounds for future large-scale neuromorphic photonic circuits. The position links to a
European collaborative project SpikePro (Spiking Photonic-Electronic IC for Quick and Efficient Processing) within the European Innovation Council framework. Close collaboration with the project partners, University of Strathclyde, University College London, TU Ilmenau and Hewlett-Packard Labs is foreseen.
The positionThis postdoctoral research position focuses on advancing neuromorphic photonic systems based on resonant tunneling diodes (RTDs). You will work closely with a PhD student who is currently developing RTD–photodetector integration for optical-to-electrical spiking in response to low-amplitude optical signals. Depending on your background and interests, you will be expected to conduct research in the following areas:
-
Increase performance and extend functionalities. Current RTD-based photonic neuron implementations are at a preliminary stage. You will contribute to significantly enhancing their performance in terms of footprint, energy efficiency, and operating speed. You will also contribute to extend the spiking building block functionalities (photodetectors, lasers, modulators) to build more complex circuits.
-
Neuromorphic event-based sensing. You will perform proof-of-concept experiments demonstrating intelligent sensing capabilities using RTD-based photonic chips. The specific sensing use case will be defined in a later stage.
The work will comprise concept development, device design (incl. layerstack, optical and electrical simulations), photonic chip layout, and chip characterization in our laboratories. The chip fabrication will be carried out by a colleague. You will be part of the Photonic Integration (PhI) research group within the Eindhoven Hendrik Casimir Institute (EHCI). The research will be done in collaboration with European partners from academia and industry.