Articolul precedent |
Articolul urmator |
261 11 |
Ultima descărcare din IBN: 2024-01-10 13:02 |
SM ISO690:2012 LUPU, Maria, KLENOV, Nikolai V.. Spintronic Functional Nanostructures for Artificial Neural Network. In: Electronics, Communications and Computing: IC ECCO 2022, Ed. 12, 20-21 octombrie 2022, Chişinău. Chișinău: Tehnica-UTM, 2023, Editia 12, pp. 34-35. |
EXPORT metadate: Google Scholar Crossref CERIF DataCite Dublin Core |
Electronics, Communications and Computing Editia 12, 2023 |
|||||||
Conferința "Electronics, Communications and Computing" 12, Chişinău, Moldova, 20-21 octombrie 2022 | |||||||
|
|||||||
Pag. 34-35 | |||||||
|
|||||||
Descarcă PDF | |||||||
Rezumat | |||||||
Energy consumption reduction becomes a crucial parameter constraining the advance of supercomputers. The non-von Neumann architectures, first of all – the Artificial Neural Networks (ANN) based on superconducting spintronic elements, seems to be the most promising solution. Superconducting ANN needs elaboration of two main elements – nonlinear one (neuron) [1] and linear connecting element (synapse) [2]. Results of our theoretical and experimental study of the proximity effect in a stack-like superconductor/ferromagnet (S/F) superlattice with Co- ferromagnetic layers of different thicknesses and coercive fields, and Nb-superconducting layers of constant thickness equal to coherence length of niobium are presented. Superconducting spin-valves and superconducting synapse, based on layered hybrid S/F nanostructures was designed and investigated. The layered nanostructures Nb/Co demonstrate change of the superconducting order parameter in thin s-films due to switching from the parallel to the antiparallel alignment of neighboring F-layers. We argue that such superlattices can be used as tunable kinetic inductors for ANN synapses design. |
|||||||
|