Variational inference (VI) is an approximation of statistically valid Bayesian inference which is well-suited for analog accelerators, and stochastic nanomagnetic devices in particular are a strong candidate to implement this feature by exploiting tunable randomness in magnetic thin-films that can be run quickly and with a low power-draw. In this work, we a) discuss how VI can be reliably implemented with a combination of low-noise nanomagnetic synapses and tunable noise generating sources (magnetic tunnel junctions (MTJs) in a single analog design and b) summarize efforts to characterize the state-dependent noise profiles of various MTJ designs for various applications.
The spin and valley physics in 2-dimensional van der Waals materials provides a unique platform for novel applications in spintronics and valleytronics. 2H phase transition metal dichalcogenides (TMD) monolayers possesses broken inversion symmetry and strong spin-orbit coupling, leading to a coupled spin and valley physics that makes them better candidates for these applications. For practical device applications, spin and valley Hall effect (SVHE) is a good way of charge to spin and charge to valley conversion, making the electrical generation of spin and valley polarization possible. While SVHE has been observed via optical measurements at cryotemperatures below 30 K, the behavior at elevated temperatures and thorough understanding of the data are still lacking. In this work we conduct spatial Kerr rotation (KR) measurements on monolayer tungsten diselenide (WSe2) field effect transistors and study the electrical control and temperature dependence of SVHE. We image the distribution of the spin and valley polarization directly and find clear evidence of the spin and valley accumulation at the edges. We show that the SVHE can be electrically modulated by the gate and drain bias, and the polarization persists at elevated temperatures. We then conduct four-port electrical test reflection spectra measurement and use a drift-diffusion model to interpret the data and extract key parameters. A lower-bound spin/valley lifetime is predicted of 40 ns and a mean free path of 240 nm below 90 K. The spin/valley polarization on the edge is calculated to be ~4% at 45 K. WSe2-on-hBN samples are prepared as well, and the KR measurements on these samples are discussed.
Magnetic tunnel junctions (MTJs) show great promise for implementation in high-performance STT-MRAM and novel computing regimes such as magnetic logic and neuromorphic computing. However, a handful of material setbacks stand in the way of the adoption of leading MgO MTJs over other emerging technologies, such as Resistive-RAM junctions, in next-generation architectures. Here, we explore the properties of iron / scandium nitride (ScN) magnetoresistive junctions using density functional theory (DFT) and find ScN a promising barrier material given its novel electron symmetry filtering properties, high TMR, and low RA-product. Magnetoresistance ratios exceeding 1900% are enabled by Δ2’ symmetry filtering through the barrier, in addition to the traditional Δ1 symmetries observed in MgO MTJs. The electronic properties of the diffusive Fe/ScN interface are resolved, with predicted half-metallicity that could amplify MR in realistic low-power ScN devices.
We propose a four-terminal domain wall-magnetic tunnel junction (DW-MTJ) neuron that enables the first-ever purely spintronic multilayer perceptron with unsupervised learning. The leaky integrate-and-fire neuron has a ferromagnetic DW track coupled to a binary MTJ by an electrically insulated layer. Current through the DW track performs integration by moving the DW. Leaking occurs by moving the DW in the opposite direction of integration due to either dipolar magnetic field, anisotropy gradient, or shape variation. When the DW passes underneath the MTJ, it fires by switching between the resistive and conductive states.
In a crossbar perceptron, the DW track of each neuron is connected to the analog three-terminal DW-MTJ synapses and the MTJ terminals cascade multiple layers. Finally, an unsupervised learning algorithm results from the feedback between the neuron MTJ and the analog synapses, providing best results of 98.11% accuracy on the Wisconsin breast cancer clustering task.
Magnetic domain-wall devices, modulated by the spin-transfer torque or the spin-orbit torque effect, can implement logical operations in a manner that is inherently compact and cascadable. Using circuit simulations with micromagnetics-validated compact models, we evaluate the device requirements for domain-wall logic that has low latency, outperforms scaled CMOS logic in energy efficiency, and remains robust to process variations. We further show how the inherent non-volatility of these devices can be leveraged to construct stateful logic circuits that save energy and area relative to their CMOS counterparts and propose novel logic architectures that exploit the unique advantages of domain-wall devices.
Neuromorphic computing captures the quintessential neural behaviors of the brain and is a promising candidate for the beyond-von Neumann computer architectures, featuring low power consumption and high parallelism. The neuronal lateral inhibition feature, closely associated with the biological receptive field, is crucial to neuronal competition in the nervous system as well as its neuromorphic hardware counterpart. The domain wall - magnetic tunnel junction (DW-MTJ) neuron is an emerging spintronic artificial neuron device exhibiting intrinsic lateral inhibition. This work discusses lateral inhibition mechanism of the DW-MTJ neuron and shows by micromagnetic simulation that lateral inhibition is efficiently enhanced by the Dzyaloshinskii-Moriya interaction (DMI).
The challenge of developing an efficient artificial neuron is impeded by the use of external CMOS circuits to perform leaking and lateral inhibition. The proposed leaky integrate-and-fire neuron based on the three terminal magnetic tunnel junction (3T-MTJ) performs integration by pushing its domain wall (DW) with spin-transfer or spin-orbit torque. The leaking capability is achieved by pushing the neurons’ DWs in the direction opposite of integration using a stray field from a hard ferromagnet or a non-uniform energy landscape resulting from shape or anisotropy variation. Firing is performed by the MTJ stack. Finally, analog lateral inhibition is achieved by dipolar field repulsive coupling from each neuron. An integrating neuron thus pushes slower neighboring neurons’ DWs in the direction opposite of integration. Applying this lateral inhibition to a ten-neuron output layer within a neuromorphic crossbar structure enables the identification of handwritten digits with 94% accuracy.
Advances in machine intelligence have sparked interest in hardware accelerators to implement these algorithms, yet embedded electronics have stringent power, area budgets, and speed requirements that may limit non- volatile memory (NVM) integration. In this context, the development of fast nanomagnetic neural networks using minimal training data is attractive. Here, we extend an inference-only proposal using the intrinsic physics of domain-wall MTJ (DW-MTJ) neurons for online learning to implement fully unsupervised pattern recognition operation, using winner-take-all networks that contain either random or plastic synapses (weights). Meanwhile, a read-out layer trains in a supervised fashion. We find our proposed design can approach state-of-the-art success on the task relative to competing memristive neural network proposals, while eliminating much of the area and energy overhead that would typically be required to build the neuronal layers with CMOS devices.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.