RESPITE

Computing with light using integrated optics has seen huge progress over the last 3-4 years in multiple fields such as neuromorphic computing, quantum computing, and on-chip data storage. This has created a vast ecosystem that relies on high-speed reconfigurations of nano-photonic circuits (such as their use as synapses or in routing applications) and ultra fast yet high-resolution, low-power photo detection. Currently, it is impossible to combine all these functionalities into an integrated platform that fits onto a single chip.

RESPITE targets a novel platform, which combines vision and cognition on a single chip. This is utilized by using novel superconducting Joule switches as neurons, multi-level phase-change memory elements as synaptic weights, and superconducting single-photon detector arrays as retina. This new platform will allow in-sensor neuromorphic computing with unprecedented performance levels. The platform will have atto-Joule switching power consumption, sub-nanosecond latency, and high compactness (3000 neurons and >100K synapses on <5 mm2). Unlike other superconducting neuromorphic technologies, the new RESPITE platform will be scalable, easy to fabricate, and compatible with low-cost cryostats, high-Tc superconductors, quantum applications, and on-chip learning architectures – making it a game changer for a wide range of users and disciplines.

Within RESPITE, Fraunhofer HHI will develop novel high-performance tiny neural networks (tNNs) adapted to the new platform. Furthermore, for these tNNs, evaluation methods for the new platform will be specified. The tNNs will be developed in 3 stages, with stage 1 only having 16 input, 6 hidden, and 3 output neurons, while stage 3 will have several thousands of neurons. For this, new technologies of value range reduction and prioritized neuron connections for efficient NN architecture sparsification will be developed.

• Single Quantum BV
• Technische Universiteit Delft
• The Chancellor, Masters and Scholars of the University of Oxford
• Universiteit Gent
• Rijksuniversiteit Groningen
• Turkiye Bilimsel ve Teknolojik Arastirma Kurumu
• Fraunhofer Gesellschaft zur Förderung der angewandten Forschung e.V.