Characterization of a network of coupled spin-torque nano-oscillators for Ising-spin based analog computing
https://www.spintec.fr/wp-content/uploads/2025/10/Plaquette_Spintec_2026.pdf
Context
Digital computers based on conventional CMOS hardware reach their limits in terms of energy consumption. Novel approaches are under investigation, such as analog computers that exploit physics concepts and reduce overall energy consumption due to their inherent parallel computing capabilities. A powerful example is based on the Ising model, where binary-valued Ising spins ππ, ππ are coupled through coupling constants Jij. Depending on the coupling matrix and operational conditions, such an Ising-based analog computing platform can be used for memory, logic, or optimization-problem-solving tasks [1]. This versatility is enhanced when exploiting the intrinsic thermal noise, thus making use of naturally available energy resources to explore the energy landscape [2].
At SPNTEC, we recently have built such an analog computing platform based on spintorque nano-oscillators, which are magnetic tunnel junctions that use spintronics concepts to convert a DC voltage into an oscillating output signal oscillating at RF frequencies. The aim of the internship is to use this Ising-model analog computing platform to demonstrate the potential for different computing tasks: optimization, logic, or memory tasks [1]. Such implementation requires as a first step a full understanding of the coupled dynamics of spintorque oscillators that will be the focus of this internship.
[1] Knoll et al. NPJ Unconv. Computing 1, 5 (2024) https://doi.org/10.1038/s44335-024-00005-1; Cai et al. Applied Phys. A 129, 236 (2023) https://doi.org/10.1007/s00339-022-06365-4
[2] N.-T. Phan et al. Phys. Rev. Appl. 21, 034063(2024) DOI: 10.1103/PhysRevApplied.21.034063
Work program & Skills acquired during internship
The internship has two options to address the scientific questions: either experimental or by simulation. It will combine fundamental studies of the phase dynamics of coupled spintronics oscillators with electrical characterization of the spintronic network, needed to control the coupling strength and phase. The student will acquire expertise in (i) concepts of spintronics (spin-polarized transport, spin momentum transfer, spin-torque driven magnetization dynamics); (ii) concepts of coupled dynamical systems; (iii) concepts of unconventional computing and (iv) high-frequency measurement techniques. The work is carried out in interaction with the different members of Spintecβs RF Spintronics and Artificial intelligence teams.
Interested students, please send a CV and a motivation letter.
- Requested background: Master 2 in condensed matter physics and/or in nanosciences; good taste for experiments or programming
- Duration: 4-6 months
- Start period: February/March 2025
- Possibility of PhD thesis : YES