Hardware security primitives that preserve secrets are playing a crucial role in the Internet-of-Things (IoT) era. Existing physical unclonable function (PUF) instantiations, exploiting static randomness, generate challenge-response pairings (CRPs) to produce unique security keys that can be used to authenticate devices linked to the IoT. Reconfigurable PUFs (RPUFs) with dynamically refreshable CRPs can enhance the security and robustness of conventional PUFs. The in-plane current-driven perpendicular polarized nanomagnet switching via spin-orbit torque (SOT) possesses great potential for application to memory and logic, as the write-current path is separate from the read-current path, which naturally resolves the write-read interference. However, the stochastic switching of perpendicular magnetization, without an additional symmetry-breaking field, would significantly hinder the technological viability of commercial implementations. Here, we report an initialization-free physical RPUF implemented by SOT-induced stochastic switching of perpendicularly magnetized Ta/CoFeB/MgO nanodevices. Using a 15 × 15 nanomagnet array, we experimentally demonstrate a security primitive that offers a near-ideal 50% uniqueness over 100 reconfiguration cycles, as well as a low correlation coefficient between every two reconfiguration cycles. Our results show that current-induced nanomagnets switching paves the way for developing highly reliable and energy-efficient reconfigurable cryptographic primitives with a smaller footprint.

Analogue arithmetic operations are the most fundamental mathematical operations used in image and signal processing as well as artificial intelligence (AI). In-memory computing offers high performance and energy-efficient computing paradigm. To date, in-memory analogue arithmetic operation with emerging nonvolatile devices were usually implemented using discrete components, which limits the scalability and blocks large scale integration. Here, we experimentally demonstrate a prototypical implementation of in-memory analogue arithmetic operations (summation, subtraction and multiplication), based on in-memory electrical current sensing units using spinorbit torque (SOT) devices. The proposed analogue arithmetic operation structures are smaller than the state-of-the-art CMOS counterparts by several orders of magnitude.Moreover, data to be processed and computing results can be locally stored, or the analogue computing can be done in the nonvolatile SOT devices, which were exploited to experimentally implement image edge detection and signal amplitude modulation with simple structure. Furthermore, we constructed an artificial neural network (ANN) with SOT devices based synapses to realize pattern recognition with high accuracy of ~95%.

No abstract

Magnetic domain wall (DW)-based logic devices offer numerous opportunities for emerging electronics applications allowing superior performance characteristics such as fast motion, high density, and nonvolatility to process information. However, these devices rely on an external magnetic field, which limits their implementation; this is particularly problematic in large-scale applications. Multiferroic systems consisting of a piezoelectric substrate coupled with ferromagnets provide a potential solution that provides the possibility of controlling magnetization through an electric field via magnetoelastic coupling. Strain-induced magnetization anisotropy tilting can influence the DW motion in a controllable way. We demonstrate a method to perform all-electrical logic operations using such a system. Ferromagnetic coupling between neighboring magnetic domains induced by the electric-field-controlled strain has been exploited to promote noncollinear spin alignment, which is used for realizing essential building blocks, including DW generation, propagation, and pinning, in all implementations of Boolean logic, which will pave the way for scalable memory-in-logic applications.