Research
Advances in engineering optical and hybrid systems allow us to realise increasingly complex physical systems. It is therefore an exciting time for theoretical physicists to devise coupled multimode systems which can be harnessed for quantum science and technological applications. In this new Max Planck Research Group, we are exploring two directions fuelled by these advances.
Topology in driven-dissipative quantum systems
Topology is a powerful principle for understanding many physically diverse complex systems and is a major research theme in condensed matter physics. Recently, a different notion of topology in systems experiencing gain and loss has been investigated, sparking the field of non-Hermitian topology, which can lead to a variety of phenomena with no Hermitian counterparts. In particular, we have shown that non-trivial non-Hermitian topology corresponds one-to-one with the phenomenon of directional amplification, meaning signals are amplified in one direction but blocked or attenuated in the other. Such unidirectionality is highly sought-after for quantum information processing applications. We recently collaborated with Ewold Verhagen´s group at AMOLF, Amsterdam, and demonstrated topological amplification in an optomechanical experiment. Furthermore, non-Hermitian topology is a resource for sensing. The group is continuing to explore how the unique properties of non-Hermitian, topological systems can be harnessed for technological applications such as signal routing, non-reciprocity and directional amplification, sensing, as well as multi-partite entanglement generation.
Further reading:
C. C. Wanjura, M. Brunelli, A. Nunnenkamp. Topological framework for directional amplification in driven-dissipative cavity arrays. Nature Communications 11, 3149 (2020).
C. C. Wanjura, M. Brunelli, A. Nunnenkamp. Correspondence between non-Hermitian topology and directional amplification in the presence of disorder. Phys. Rev. Lett. 127, 213601 (2021).
M. Brunelli, C. C. Wanjura, A. Nunnenkamp. Restoration of the non-Hermitian bulk-boundary correspondence via topological amplification. SciPost Phys. 15, 173 (2023).
C. C. Wanjura, J. J. Slim, J. del Pino, M. Brunelli, E. Verhagen, A. Nunnenkamp. Quadrature nonreciprocity in bosonic networks without breaking time-reversal symmetry. Nature Physics 19, 1429–1436 (2023).
J. J. Slim, C. C. Wanjura, M. Brunelli, J. del Pino, A. Nunnenkamp, E. Verhagen. Optomechanical realization of the bosonic Kitaev chain. Nature 627, 767–771 (2024).
You can learn more on non-Hermitian physics in this online seminar series.
Neuromorphic Computing
With the trend of exponentially growing digital neural network sizes and the associated energy consumption of machine learning applications, there is a need for alternative hardware approaches. In response to this demand, neuromorphic computing aims to replace our digital neural networks with physical systems. Indeed, optical or photonic systems can be engineered to perform the desired machine learning tasks and are a promising platform for neuromorphic computing as they offer high computation speeds at low energy consumption. However, realizing the necessary non-linear computation is challenging. We made an important step towards overcoming this challenge by developing a new framework for performing non-linear computation with purely linear scattering (check out our video below).
This approach can be implemented in essentially any linear system with access to a sufficient number of tunable parameters, in particular, in existing scalable state-of-the-art platforms such as optics, microwave and electrical circuits and integrated photonics. Another very important aspect is physical training, namely physical access to gradients. It has been shown that in-silico training, i.e. training in simulation, does not generally transfer well to experiment since even small discrepancies between model and reality can lead to a disastrous accumulation of errors during training. Therefore, we are looking into new training approaches for neuromorphic systems.
Learn more on non-linear computation with linear wave scattering in this video:
Contact
Research Group Clara Wanjura
Max Planck Institute for the Science of Light
Staudtstr. 2
91058 Erlangen, Germany
