SKM 2023 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
MA: Fachverband Magnetismus
MA 29: Neuromorphic Magnetism / Magnetic Logic
MA 29.2: Vortrag
Mittwoch, 29. März 2023, 11:45–12:00, HSZ 02
Superparamagnetic tunnel junctions for neuromorphic computing — •Leo Schnitzspan1,2, Gerhard Jakob1,2, and Mathias Kläui1,2 — 1Institut für Physik, Johannes Gutenberg Universität Mainz — 2Max Planck Graduate Center, Mainz
Superparamagnetic tunnel junctions (SMTJ) are promising building blocks in the field of neuromorphic computing. In a SMTJ, the magnetic free layer can switch its magnetic orientation induced by thermal activation, leading to a random two-level resistance fluctuations[1]. We show nanosecond fluctuations with dwell times below 10 ns for our in-plane magnetized SMTJs. Their intrinsic stochastic behaviour and additional tunability by external magnetic fields, Spin Transfer Torques (STT) or Spin Orbit Torques (SOT) are prerequisites for low-energy artificial neurons in neural networks. True random number generation is demonstrated and evaluated by the statistical test suite from NIST. The probability of a P- (=0) or AP- (=1) state depends on the energy landscape and can be affected by STT. However, the average fluctuation speed is strongly dependend on the temperature. We demonstrate that Joule heating, induced by a large applied current, leads to significantly shorter dwell times. From dwell time measurements, the contributions of STT and Joule heating are extracted.
[1] Hayakawa, K. et al., Phys. Rev. Lett. 126, 117202 (2021).