European Training Network on Post-Digital Computing (Post-Digital)

Informazioni relative al progetto

POST-DIGITAL

ID dell’accordo di sovvenzione: 860360

DOI

10.3030/860360

Progetto chiuso

Data della firma CE 27 Marzo 2020

Data di avvio 1 Aprile 2020

Data di completamento 31 Marzo 2024

Finanziato da

• EXCELLENT SCIENCE - Marie Skłodowska-Curie Actions

Costo totale

€ 4 021 262,28

Contributo UE

€ 4 021 262,28

Coordinato da

ASTON UNIVERSITY
United Kingdom

CORDIS fornisce collegamenti ai risultati finali pubblici e alle pubblicazioni dei progetti ORIZZONTE.

I link ai risultati e alle pubblicazioni dei progetti del 7° PQ, così come i link ad alcuni tipi di risultati specifici come dataset e software, sono recuperati dinamicamente da .OpenAIRE .

Risultati finali

Documents, reports (23)

Report on final prototype of integrated neuromorphic system with integrated readout
Network applications of neuromorphic computing (OTAW3)
3rd year POST-DIGITAL workshop

3rd year POST-DIGITAL workshop

Report on the design of photonic- and electronic- based analogue computing

Report on the design of photonic- and electronic- based analogue computing.

Report on first prototype of integrated neuromorphic system with integrated readout
Open-to-all Workshop I (OTAW I) (details on courses provided and ESR attendance record)
Final report on publications and public engagement activities completed
POST-DIGITAL Mini-symposia: Progress online adaptation of conceptor-controlled neural networks
Open-to-all Workshop II (OTAW II) (details on courses provided and ESR attendance record)
Simulation-based evaluation of multi-reservoir design methodology for standard and advanced benchmarks
1st year POST-DIGITAL workshop and TSW I (details on courses provided)
POST-DIGITAL Mini-symposia: Design methodology for powerful hierarchical integrated analog computing systems
2nd year POST-DIGITAL workshop (details on courses provided and ESR attendance record)
Report on the characterization of the photonic- and electronic- based analogue computing implementations
Report on comparison of neuromorphic computation solutions, with respect to performance on standard tasks, speed, energy consumption and footprint.
Report on critical gaps in existing formal methods to capture computation in unconventional substrates
Final POST-DIGITAL workshop (details on courses provided and ESR attendance record)
Report on novel approaches (I) core formal ideas and demo implementations in simulation and (II) worked-out formalism and demo implementations in unconventional hardware
Interim report on publications and public engagement activities completed
"International Summer School on the topic of ""Novel Nonlinear Substrates for Neural Networks"""

International Summer School on the topic of “Novel Nonlinear Substrates for Neural Networks”

Simulation and experimental study of application of photonic reservoir computing to non-linear optical channel equalization
Report on the evaluation of the photonic- and electronic- based analogue computing implementations
POST-DIGITAL Mini-symposia: Integrated photonic reservoir computing for telecom applications

POST-DIGITAL Mini-symposia: Integrated photonic reservoir computing for telecom applications

Other (1)

15 Doctoral degrees awarded

Websites, patent fillings, videos etc. (1)

POST-DIGITAL website completed

POST-DIGITAL website completed

Pubblicazioni

Peer reviewed articles (29)

Hands-on reservoir computing: a tutorial for practical implementation

Autori: Matteo Cucchi, Steven Abreu, Giuseppe Ciccone, Daniel Brunner and Hans Kleemann
Pubblicato in: Neuromorphic Computing and Engineering, Numero Volume 2, Number 3, 2022, ISSN 2634-4386
Editore: IOP Publishing Ltd
DOI: 10.1088/2634-4386/ac7db7

Computational metrics and parameters of an injection-locked large area semiconductor laser for neural network computing

Autori: Anas Skalli, Xavier Porte, Nasibeh Haghighi, Stephan Reitzenstein, James A. Lott, and Daniel Brunner
Pubblicato in: Optical Materials Express, Numero 21593930, 2022, ISSN 2159-3930
Editore: Optical Society of America
DOI: 10.1364/ome.451524

Integrated programmable spectral filter for frequency-multiplexed neuromorphic computers

Autori: TIGERS JONUZI, ALESSANDRO LUPO, MIGUEL C.SORIANO, SERGE MASSAR, J. D. DOMENÉCH
Pubblicato in: Optics Express, Numero 31(12), 2023, Pagina/e 19255-19265, ISSN 1094-4087
Editore: Optical Society of America
DOI: 10.1364/oe.489246

A complete, parallel and autonomous photonic neural network in a semiconductor multimode laser

Autori: Xavier Porte, Anas Skalli, Nasibeh Haghighi, Stephan Reitzenstein, James A Lott, Daniel Brunner
Pubblicato in: Journal of Physics: Photonics, Numero 3/2, 2021, Pagina/e 024017, ISSN 2515-7647
Editore: IOP Institute of Physics
DOI: 10.1088/2515-7647/abf6bd

Exploiting oscillatory dynamics of delay systems for reservoir computing

Autori: Mirko Goldmann; Ingo Fischer; Claudio R. Mirasso; Miguel C. Soriano
Pubblicato in: Chaos, Numero 33, 2023, Pagina/e 093139, ISSN 1089-7682
Editore: AIP Publishing
DOI: 10.1063/5.0156494

56 GBaud PAM-4 100 Km Transmission System With Photonic Processing Schemes

Autori: Irene Estébanez; Shi Li; Janek Schwind; Ingo Fischer; Stephan Pachnicke; Apostolos Argyris
Pubblicato in: Journal of Lightwave Technology, 2021, Pagina/e 55 - 62, ISSN 0733-8724
Editore: Optical Society of America
DOI: 10.1109/jlt.2021.3117921

Photonic reservoir computing for nonlinear equalization of 64-QAM signals with a Kramers–Kronig receiver

Autori: Sarah Masaad, Emmanuel Gooskens , Stijn Sackesyn , Joni Dambre and Peter Bienstman
Pubblicato in: Nanophotonics, 2022, ISSN 2192-8606
Editore: De Gruyter
DOI: 10.1515/nanoph-2022-0426

High Speed Human Action Recognition using a Photonic Reservoir Computer

Autori: Enrico Picco, Piotr Antonik, Serge Massar
Pubblicato in: Neural Networks,, Numero Volume 165,, 2023, Pagina/e 662-675, ISSN 0893-6080
Editore: Pergamon Press Ltd.

Weak Kerr Nonlinearity Boosts the Performance of Frequency-Multiplexed Photonic Extreme Learning Machines: A Multifaceted Approach

Autori: Marina Zajnulina, Alessandro Lupo, Serge Massar
Pubblicato in: Optics Express, 2024, ISSN 1094-4087
Editore: Optical Society of America
DOI: 10.1364/oe.503279

Photonic neuromorphic computing using vertical cavity semiconductor lasers

Autori: Anas Skalli, Joshua Robertson, Dafydd Owen-Newns, Matej Hejda, Xavier Porte, Stephan Reitzenstein, Antonio Hurtado, D. Brunner
Pubblicato in: Optical Materials Express, Numero 12, 2022, Pagina/e 2395-2414, ISSN 2159-3930
Editore: Optical Society of America
DOI: 10.1364/ome.450926

Experimental implementation of a neural network optical channel equalizer in restricted hardware using pruning and quantization

Autori: Diego Argüello Ron, Pedro J. Freire, Jaroslaw E. Prilepsky, Morteza Kamalian-Kopae, Antonio Napoli, Sergei K. Turitsyn
Pubblicato in: Scientific Reports, Numero 12, 2022, ISSN 2045-2322
Editore: Nature Publishing Group
DOI: 10.1038/s41598-022-12563-0

Information Processing Capacity of a Single-Node Reservoir Computer: An Experimental Evaluation

Autori: Benedikt Vettelschoss; André Röhm; Miguel C. Soriano
Pubblicato in: IEEE Transactions on Neural Networks and Learning Systems, Numero Special Numero on New Frontiers in Extremely Efficient Reservoir Computing, 2021, ISSN 2162-237X
Editore: IEEE Computational Intelligence Society
DOI: 10.1109/tnnls.2021.3116709

Online digital compensation of pump dithering induced phase and amplitude distortions in transmission links with cascaded fibre-optical parametric amplifiers

Autori: Long H. Nguyen, Sonia Boscolo, and Stylianos Sygletos
Pubblicato in: Opt. Express, Numero 32, 2024, Pagina/e 13467-13477, ISSN 1094-4087
Editore: Optical Society of America
DOI: 10.1364/oe.516840

Photonic extreme learning machine based on frequency multiplexing

Autori: Alessandro Lupo, Lorenz Butschek, and Serge Massar
Pubblicato in: Optics Express, Numero Volume . 29,Numero 18,, 2021, Pagina/e p. 28257-28276, ISSN 1094-4087
Editore: Optical Society of America
DOI: 10.1364/oe.433535

Learn one size to infer all: Exploiting translational symmetries in delay-dynamical and spatio-temporal systems using scalable neural networks

Autori: Mirko Goldmann, Claudio R. Mirasso, Ingo Fischer, Miguel C. Soriano
Pubblicato in: PHYSICAL REVIEW E, Numero Vol. 106, Iss. 4, 2022, ISSN 2470-0045
Editore: American Physical Society
DOI: 10.48550/arxiv.2111.03706

Photonic reservoir computer based on frequency multiplexing

Autori: Lorenz Butschek, Akram Akrout, Evangelia Dimitriadou, Alessandro Lupo, Marc Haelterman, Serge Massar
Pubblicato in: Optics Letters, Numero Vol. 47, Numero 4,, 2022, Pagina/e pp. 782-785, ISSN 0146-9592
Editore: Optical Society of America
DOI: 10.1364/ol.451087

A photonics perspective on computing with physical substrates

Autori: S. Abreu, I. Boikov, M. Goldmann, T. Jonuzi, A. Lupo, S. Masaad, L. Nguyen, E. Picco, G. Pourcel, A. Skalli, L. Talandier, B. Vettelschoss, E.A. Vlieg, A. Argyris, P. Bienstman, D. Brunner, J. Dambre, L. Daudet, J.D. Domenech, I. Fischer, F. Horst, S. Massar, C.R. Mirasso, B.J. Offrein, A. Rossi, M.C. Soriano, S. Sygletos, S.K. Turitsyn
Pubblicato in: Reviews in Physics, Numero 12, 2024, Pagina/e 100093, ISSN 2405-4283
Editore: Elsevier B.V.
DOI: 10.1016/j.revip.2024.100093

Deep photonic reservoir computer based on frequency multiplexing with fully analog connection between layers

Autori: Alessandro Lupo, Enrico Picco, Marina Zajnulina, and Serge Massar
Pubblicato in: Optica, Numero Vol. 10, Numero 11,, 2023, Pagina/e 1478-1485, ISSN 2334-2536
Editore: OPTICA Publishing Group
DOI: 10.1364/optica.489501

Photonic neuromorphic technologies in optical communications

Autori: Apostolos Argyris
Pubblicato in: Nanophotonics, Numero Volume 11 Numero 5, 2022, ISSN 2192-8606
Editore: De Gruyter
DOI: 10.1515/nanoph-2021-0578

Parallel Extreme Learning Machines Based on Frequency Multiplexing

Autori: Alessandro Lupo, Serge Massar
Pubblicato in: MDPI Applied Sciences, Numero Volume 12, Numero 1, Special Numero Neuromorphic Photonics: Current Devices, Systems and Perspectives, 2021, Pagina/e p. 214, ISSN 2076-3417
Editore: MDPI
DOI: 10.3390/app12010214

Is the Number of Trainable Parameters All That Actually Matters?

Autori: Chatelain, Amélie; Djeghri, Amine; Hesslow, Daniel; Launay, Julien; Poli, Iacopo
Pubblicato in: Proceedings of Machine Learning Research, Numero 163, 2022, ISSN 2640-3498
Editore: MLR Press

Exploiting transient dynamics of a time-multiplexed reservoir to boost the system performance

Autori: Mirko Goldmann, Claudio R. Mirasso, Ingo Fischer, Miguel C. Soriano
Pubblicato in: 2021 International Joint Conference on Neural Networks (IJCNN), 2021, ISSN 2161-4407, ISBN 978-1-6654-3900-8
Editore: IEEE
DOI: 10.1109/ijcnn52387.2021.9534333

Nonlinear photonic dynamical systems for unconventional computing

Autori: Daniel Brunner; Laurent Larger; Miguel Cornelles Soriano
Pubblicato in: Nonlinear Theory and Its Applications, IEICE, Numero 21854106, 2022, ISSN 2185-4106
Editore: jstage
DOI: 10.1587/nolta.13.26

Boolean learning under noise-perturbations in hardware neural networks

Autori: Louis Andreoli, Xavier Porte, Stéphane Chrétien, Maxime Jacquot, Laurent Larger, Daniel Brunner
Pubblicato in: Nanophotonics, Numero 9/13, 2020, Pagina/e 4139-4147, ISSN 2192-8614
Editore: De Gruyter
DOI: 10.1515/nanoph-2020-0171

Evanescent coupling of nonlinear integrated cavities for all-optical reservoir computing

Autori: I K Boikov, D Brunner, A De Rossi
Pubblicato in: New Journal of Physics, Numero 25, 2023, Pagina/e 093056, ISSN 1367-2630
Editore: Institute of Physics Publishing
DOI: 10.1088/1367-2630/acfba6

Analog information processing with time-multiplexed optoelectronic systems

Autori: Mirko Goldmann, Ingo Fischer, Miguel C. Cornelles-Soriano
Pubblicato in: Emerging Topics in Artificial Intelligence (ETAI) 2021, Numero 11804, 118041X, 2022, ISSN 0277-786X
Editore: SPIE
DOI: 10.1117/12.2594151

Reducing Computational Complexity of Neural Networks in Optical Channel Equalization: From Concepts to Implementation

Autori: Pedro J. Freire, Antonio Napoli, Bernhard Spinnler, Michael Anderson, Diego Argüello Ron, Wolfgang Schairer, Thomas Bex, Nelson Costa, Sergei K. Turitsyn, Jaroslaw E. Prilepsky
Pubblicato in: Journal of Lightwave Technology, Numero 41, 2024, Pagina/e 4557-4581, ISSN 0733-8724
Editore: Optical Society of America
DOI: 10.1109/jlt.2023.3234327

Toward a formal theory for computing machines made out of whatever physics offers

Autori: Herbert Jaeger, Beatriz Noheda & Wilfred G. van der Wiel
Pubblicato in: Nature Communications, Numero volume 14, Article number: 4911 (2023), 2023, ISSN 2041-1723
Editore: Nature Publishing Group

An Integrated Photorefractive Analog Matrix-Vector Multiplier for Machine Learning

Autori: Elger A. Vlieg,Lucas Talandier, Roger Dangel, Folkert Horst, Bert J. Offrein
Pubblicato in: Applied Sciences, Numero 12(9), 2022, Pagina/e 4226, ISSN 2076-3417
Editore: MDPI
DOI: 10.3390/app12094226

Conference proceedings (9)

The Demonstration of Photorefractive Synaptic Connections for an Integrated Photonic Crossbar Array

Autori: Elger A. Vlieg, Folkert Horst, Roger Dangel, Bert J. Offrein
Pubblicato in: 2023 Conference on Lasers and Electro-Optics Europe & European Quantum Electronics Conference (CLEO/Europe-EQEC), 2023, Pagina/e 1-1
Editore: IEEE
DOI: 10.1109/cleo/europe-eqec57999.2023.10232391

Flow cytometry with event-based vision and spiking neuromorphic hardware

Autori: Steven Abreu, Muhammed Gouda, Alessio Lugnan, Peter Bienstman
Pubblicato in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, Numero pp. 4139-4147, 2023
Editore: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops

A complete, parallel, and autonomous photonic neural network in a semiconductor multimode laser

Autori: Anas Skalli, Javier Porte, Nasibeh Haghighi, Stephan Reitzenstein, James A. Lott, Daniel Brunner
Pubblicato in: Emerging Topics in Artificial Intelligence (ETAI) 2021, 2021
Editore: SPIE
DOI: 10.1117/12.2594646

Photonic co-processors in HPC: Using LightOn OPUs for Randomized Numerical Linear Algebra

Autori: Daniel Hesslow, Alessandro Cappelli, Igor Carron, Laurent Daudet, Raphael Lafargue, Kilian Muller, Ruben Ohana, Gustave Pariente, Iacopo Poli
Pubblicato in: 2021 IEEE Hot Chips 33 Symposium (HCS), 2022, Pagina/e 1-9
Editore: IEEE
DOI: 10.1109/hcs52781.2021.9566948

What Language Model Architecture and Pretraining Objective Work Best for Zero-Shot Generalization?

Autori: Thomas Wang, Adam Roberts, Daniel Hesslow, Teven Le Scao, Hyung Won Chung, Iz Beltagy, Julien Launay, Colin Raffel
Pubblicato in: Proceedings of the 39 th International Conference on Machine Learning, 2022, ISSN 2640-3498
Editore: MLR press

Programmable photorefractive optical synapses in integrated photonics for neuromorphic computing

Autori: Elger A. Vlieg, Folkert Horst, Roger Dangel, and Bert J. Offrein
Pubblicato in: European Conference on Integrated Optics, 2023
Editore: University of Twente

Neural Network Computing using a Semiconductor Multimode Laser

Autori: Xavier Porte, Anas Skalli, Nasibeh Haghighi, Stephan Reitzenstein, James A. Lott, Daniel Brunner
Pubblicato in: OSA Nonlinear Optics 2021, 2022
Editore: Optica Publishing Group
DOI: 10.1364/nlo.2021.nth1a.2

Photonic photorefractive MAC accelerator for Machine Learning

Autori: Elger A. Vlieg, Roger Dangel, Folkert Horst, Bert J. Offrein
Pubblicato in: Photonics in Switching and Computing 2021, 2022
Editore: Optica Publishing Group
DOI: 10.1364/psc.2021.m2b.3

Motivation and Challenges for Applying Photonic Neuromorphic Computing Technologies

Autori: B.J. Offrein, E.A. Vlieg, F. Hermann, L. Carraria Martinotti, F. Horst
Pubblicato in: 2023 Conference on Lasers and Electro-Optics Europe & European Quantum Electronics Conference (CLEO/Europe-EQEC), 2023
Editore: IEEE
DOI: 10.1109/cleo/europe-eqec57999.2023.10232665

Other (13)

Concepts and Paradigms for Neuromorphic Programming

Autori: Steven Abreu
Pubblicato in: 2023
Editore: arxiv

Toward a formal theory for computing machines made out of whatever physics offers: extended version

Autori: Herbert Jaeger, Beatriz Noheda, Wilfred G. van der Wiel
Pubblicato in: 2023
Editore: ArXive
DOI: 10.48550/arxiv.2307.15408

Nonlinear integrated optical resonators for optical fibre data recovery

Autori: Ivan K. Boikov, Daniel Brunner, Alfredo De Rossi
Pubblicato in: 2024
Editore: ArXiv

Neuromorphic Intermediate Representation: A Unified Instruction Set for Interoperable Brain-Inspired Computing

Autori: Pedersen, Jens E.; Abreu, Steven; Jobst, Matthias; Lenz, Gregor; Fra, Vittorio; Bauer, Felix C.; Muir, Dylan R.; Zhou, Peng; Vogginger, Bernhard; Heckel, Kade; Urgese, Gianvito; Shankar, Sadasivan; Stewart, Terrence C.; Eshraghian, Jason K.; Sheik, Sadique
Pubblicato in: 2023
Editore: arXive
DOI: 10.48550/arxiv.2311.14641

Training a spiking neural network on an event-based label-free flow cytometry dataset

Autori: Gouda, Muhammed; Abreu, Steven; Lugnan, Alessio; Bienstman, Peter
Pubblicato in: arxiv, Numero 33, 2023
Editore: Arxiv
DOI: 10.48550/arxiv.2303.10632

Fully analog photonic deep Reservoir Computer based on frequency multiplexing.

Autori: A. Lupo, E. Picco, M. Zajnulina, S. Massar.
Pubblicato in: Arxiv, 2023
Editore: Arxiv
DOI: 10.48550/arxiv.2305.08892

Realization of Neural Network-based Optical Channel Equalizer in Restricted Hardware

Autori: Diego R. Arguello, Pedro J. Freire, Jaroslaw E. Prilepsky, Antonio Napoli, Morteza Kamalian-Kopae, Sergei K. Turitsyn
Pubblicato in: 2021
Editore: Arxiv

Direct coupling of nonlinear integrated cavities for all-optical reservoir computing

Autori: Ivan Boikov, Daniel Brunner, Alfredo De Rossi
Pubblicato in: 2023
Editore: ArXiv

Photonic processing for a 56 GBaud PAM-4 100 km Transmission System

Autori: Estébanez, Irene; Li, Shi; Schwind, Janek; Fischer, Ingo; Pachnicke, Stephan; Argyris, Apostolos
Pubblicato in: 2022
Editore: IFISC

Contrastive Embeddings for Neural Architectures

Autori: Daniel Hesslow & Iacopo Poli
Pubblicato in: 2021
Editore: Arxiv
DOI: 10.48550/arxiv.2102.04208

Efficient Optimisation of Physical Reservoir Computers using only a Delayed Input

Autori: Picco, Enrico; Jaurigue, Lina; Lüdge, Kathy; Massar, Serge
Pubblicato in: 2023
Editore: ArXiv
DOI: 10.48550/arxiv.2401.14371

Timescales: the choreography of classical and unconventional computing

Autori: Herbert Jaeger*, Francky Catthoor
Pubblicato in: 2023
Editore: arXiv
DOI: 10.48550/arxiv.2301.00893

Deep Photonic Reservoir Computer for Speech Recognition

Autori: Enrico Picco, Alessandro Lupo, Serge Massar
Pubblicato in: 2023
Editore: Arxiv
DOI: 10.48550/arxiv.2312.06558

Diritti di proprietà intellettuale

Patent (1)

METHOD AND SYSTEM FOR MACHINE LEARNING USING OPTICAL DATA

Numero candidatura/pubblicazione: 21 305295
Data: 2021-03-11
Candidato/i: LIGHT ON