BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:IMX Talks - A magnonic matrix device for parallel convolutional pr
 ocessing
DTSTART:20260915T154500
DTEND:20260915T164500
DTSTAMP:20260916T011032Z
UID:5c6b97bf910f6421288678071015f2dea9b33c4e01c9cacd12456b85
CATEGORIES:Conferences - Seminars
DESCRIPTION:Prof. Sebastiaan van Dijken\, Department of Applied Physics A
 alto University\, Finland\nThe growing computational demands of artificial
  intelligence are exposing fundamental limits of electronic and photonic a
 ccelerators\, spurring a search for alternative physical computing hardwar
 e capable of efficient matrix-vector multiplication (MVM). Magnonic system
 s\, which encode information in propagating spin waves\, offer low-power s
 ignal transport and wavelengths well suited to dense on-chip integration\,
  yet their potential for parallel computing remains largely unexplored. He
 re we present magnonic matrix devices\, built from low-loss yttrium iron g
 arnet (YIG) waveguide networks\, that exploit spin-wave transport and inte
 rference to perform real-time\, parallel MVM and convolutional operations.
  These devices achieve high area efficiencies\, comparable to state-of-the
 -art electronic physical computing platforms\, alongside competitive power
  efficiencies. Integrated into convolutional neural networks\, they perfor
 m benchmark tasks including handwritten digit recognition (97.1% accuracy)
 \, MRI image classification (83.2%)\, and Parkinsonian gait detection (90.
 6%). Weights can be tuned globally through microwave frequency and bias ma
 gnetic field\, enabling network-wide reconfiguration for task switching\, 
 while local optical and electrical control is demonstrated as a proof of c
 oncept\, pointing toward spatially selective weight updates for future on-
 chip training. Together\, these results establish magnonic convolutional p
 rocessing as a scalable platform for efficient AI acceleration and open a 
 route toward compact\, wave-based processors that intrinsically support pa
 rallel computation.\n\nBio: Sebastiaan van Dijken is a professor in the D
 epartment of Applied Physics at Aalto University in Finland\, where he lea
 ds the Nanomagnetism and Spintronics group. His research focuses on spin-w
 ave dynamics and magnon transport\, with the goal of exploiting magnons fo
 r energy-efficient signal processing and data transfer in next-generation 
 computing systems. By designing and fabricating engineered magnetic thin f
 ilms and nanostructures\, his work advances both fundamental knowledge of 
 spin transport and magnetization dynamics and enables practical implementa
 tions in magnonic circuits and spintronic devices.\nHis group also investi
 gates complementary directions\, including electric-field control of magne
 tism and magnetoplasmonics. More recently\, his research has expanded into
  neuromorphic and in-sensor computing\, exploring how magnetic and magnoni
 c systems\, as well as photomemristor networks\, can perform tasks such as
  pattern recognition and adaptive learning\, offering new architectures in
 spired by biological neural systems.\n 
LOCATION:BM 3241 https://plan.epfl.ch/?room==BM%203241
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR