Search

The transistor celebrated its 75${}^\text{th}$ birthday in 2022. The continued scaling of the transistor defined by Moore's Law continues, albeit at a slower pace. Meanwhile, computing demands and energy consumption required by modern artificial intelligence (AI) algorithms have skyrocketed. As an alternative to scaling transistors for general-purpose computing, the integration of transistors with unconventional technologies has emerged as a promising path for domain-specific computing. In this article, we provide a full-stack review of probabilistic computing with p-bits as a representative example of the energy-efficient and domain-specific computing movement. We argue that p-bits could be used to build energy-efficient probabilistic systems, tailored for probabilistic algorithms and applications. From hardware, architecture, and algorithmic perspectives, we outline the main applications of probabilistic computers ranging from probabilistic machine learning and AI to combinatorial optimization and quantum simulation. Combining emerging nanodevices with the existing CMOS ecosystem will lead to probabilistic computers with orders of magnitude improvements in energy efficiency and probabilistic sampling, potentially unlocking previously unexplored regimes for powerful probabilistic algorithms.

INTRODUCTION: Incorporation of a plant-based diet was effective in both induction and short-term relapse prevention in Crohn’s disease. Ten-year long-term relapse-free rates in Crohn’s disease are around 10% to 23%. OBJECTIVE: We investigated whether infliximab and plant-based diet as first-line therapy enhance the long-term relapse-free rate in patients with Crohn’s disease. METHODS: This single-group, prospective study was performed in tertiary hospitals in Japan. Remission was induced in 24 consecutive newly diagnosed adult patients with Crohn’s disease during hospitalization via 3 standard infliximab infusions together with a plant-based diet. Patients were instructed to continue the diet after discharge. Scheduled maintenance infliximab infusion was not used. The primary endpoint was relapse, which was defined as the appearance of symptoms resulting in the alteration of therapeutic modality. The secondary endpoints were C-reactive protein level, plant-based diet score, and surgery. RESULTS: The median follow-up period was 8.6 years. Thirteen cases were relapse-free. The relapse-free rate evaluated by Kaplan-Meier survival analysis at 1, 2, 3, and 4 years was 79%, 66%, 57%, and 52%, respectively. There was no further reduction afterward up to 10 years. The relapse-free rate with normal C-reactive protein levels at 1 to 2 and 3 to 10 years was 57% and 52%, respectively. The plant-based diet score at 20 months and 5 years was significantly higher relative to baseline (p < 0.0001). Surgical rates at 5 and 10 years were 12% and 19%, respectively. CONCLUSIONS: Infliximab and plant-based diet as first-line therapy created an unprecedented relapse-free course in nearly half of patients with Crohn’s disease.

Modulation of the energy landscape by external perturbations governs various thermally-activated phenomena, described by the Arrhenius law. Thermal fluctuation of nanoscale magnetic tunnel junctions with spin-transfer torque (STT) shows promise for unconventional computing, whereas its rigorous representation, based on the Néel-Arrhenius law, has been controversial. In particular, the exponents for thermally-activated switching rate therein, have been inaccessible with conventional thermally-stable nanomagnets with decade-long retention time. Here we approach the Néel-Arrhenius law with STT utilising superparamagnetic tunnel junctions that have high sensitivity to external perturbations and determine the exponents through several independent measurements including homodyne-detected ferromagnetic resonance, nanosecond STT switching, and random telegraph noise. Furthermore, we show that the results are comprehensively described by a concept of local bifurcation observed in various physical systems. The findings demonstrate the capability of superparamagnetic tunnel junction as a useful tester for statistical physics as well as sophisticated engineering of probabilistic computing hardware with a rigorous mathematical foundation.

Electrical manipulation of magnetic materials by current-induced spin torque constitutes the basis of spintronics. Here, we show an unconventional response to spin–orbit torque of a non-collinear antiferromagnet Mn3Sn, which has attracted attention owing to its large anomalous Hall effect despite a vanishingly small net magnetization. In epitaxial heavy-metal/Mn3Sn heterostructures, we observe a characteristic fluctuation of the Hall resistance under the application of electric current. This observation is explained by a rotation of the chiral-spin structure of Mn3Sn driven by spin–orbit torque. We find that the variation of the magnitude of anomalous Hall effect fluctuation with sample size correlates with the number of magnetic domains in the Mn3Sn layer. In addition, the dependence of the critical current on Mn3Sn layer thickness reveals that spin–orbit torque generated by small current densities, below 20 MA cm−2, effectively acts on the chiral-spin structure even in Mn3Sn layers that are thicker than 20 nm. The results provide additional pathways for electrical manipulation of magnetic materials. Current-induced rotation in epitaxial films of the non-collinear antiferromagnet Mn3Sn is investigated.

The development of new functional memories using emerging nonvolatile devices has been widely investigated. Spin-transfer torque magnetoresistive random access memory (STT-MRAM) has become new technology platform to overcome the issue in power consumption of logic for the application from IoT to AI; however, STT-MRAM has a tradeoff relationship between endurance, retention, and access time. This is because the MTJ device used in STT-MRAM is a two-terminal device, and excessive read current for high-speed readout can cause unexpected data writing, or so-called read disturbance. In order to meet the demand for the realization of high-speed nonvolatile memory, the development of new memories based on innovative circuit, device, and integration process is required. In this article, we demonstrate an SOT-MRAM, a nonvolatile memory using MTJ devices with spin-orbit-torque (SOT) switching that have a read-disturbance-free characteristic. The SOT-MRAM fabricated using a 55-nm CMOS process is implemented in a dual-port configuration utilizing a three-terminal structure of the device for realizing a wide bandwidth applicable to high-speed applications. In addition, a read-energy reduction technique called a self-termination scheme is also implemented. Through the measurement results of the fabricated prototype chip, we will demonstrate the proposed SOT-MRAM achieves 60-MHz write and 90-MHz read operations with 1.2-V supply voltage under a magnetic-field-free condition.

Spin defect centers with long quantum coherence times ($T_2$) are key solid-state platforms for a variety of quantum applications. Recently, cluster correlation expansion (CCE) techniques have emerged as a powerful tool to simulate the $T_2$ of defect electron spins in these solid-state systems with good accuracy. Here, based on CCE, we uncover an algebraic expression for $T_2$ generalized for host compounds with dilute nuclear spin baths, which enables a quantitative and comprehensive materials exploration with a near instantaneous estimate of the coherence. We investigate more than 12,000 host compounds at natural isotopic abundance, and find that silicon carbide (SiC), a prominent widegap semiconductor for quantum applications, possesses the longest coherence times among widegap non-chalcogenides. In addition, more than 700 chalcogenides are shown to possess a longer $T_2$ than SiC. We suggest new potential host compounds with promisingly long $T_2$ up to 47 ms, and pave the way to explore unprecedented functional materials for quantum applications., 24 pages, 4 figures, 1 table

The use of high‑energy x‑ray diffraction techniques with the latest generation synchrotron sources has created new approaches to study quantitatively the structure of noncrystalline materials. Recently, this technique has been combined with neutron diffraction at pulsed source to provide more detailed and reliable structural information not previously available. This article reviews and summarises recent results obtained from the high energy x‑ray diffraction on several oxide glasses, SiO2, B2O3 and PbSiO3, using bending magnet beamlines at SPring‑8. In particular, it addresses the structural models of the oxide glasses obtained by the reverse Monte Carlo (RMC) modelling technique using both the high energy x‑ray and neutron diffraction data.

D019-Mn3Sn, an antiferromagnet having a non-collinear spin structure in a kagome lattice, has attracted great attention owing to various intriguing properties such as large anomalous Hall effect. Stability of a magnetic state against thermal fluctuation, characterized in general by the thermal stability factor Δ, has been well studied in ferromagnetic systems but not for antiferromagnets. Here, we study Δ of the antiferromagnetic Mn3Sn nanodots as a function of their diameter D. To quantify Δ, we measure the switching probability as a function of the pulse-field amplitude and analyze the results based on a model taking account of two and sixfold magnetic anisotropies in the kagome plane. We observe no significant change in Δ down to D = 300 nm below which it decreases with D. The obtained D dependence is well explained by a single-domain and nucleation-mediated reversal models. These findings provide a basis to understand the thermal fluctuation and reversal mechanism of antiferromagnets for device applications.

The demand for energy-efficient, high-performance microcontroller units (MCUs) for the use in power-supply-critical Internet-of-Things (IoT) sensor-node applications has witnessed a substantial increase. In response, research concerning the development of several low-power-consuming MCUs has been actively pursued. The performance level of such MCUs, however, has not been sufficient, thereby rendering them non-feasible for the use in IoT sensor-node applications that process a large number of received signals immediately followed by extraction of valuable information from them to limit data transferred to a data center. To realize next-generation IoT systems based on intelligent sensor-node application, ultra-low-power high-performance MCUs need to be developed. This paper presents an ultra-low-power-consuming and high-performance MCU configuration based on the spintronics device technology, using which all modules are non-volatilized, and any wasteful power consumption is eliminated by controlling the power supplied independently to each module. By incorporating a reconfigurable accelerator module, for performing various signal-processing procedures in sensor-node applications, and a memory controller, which can speed up the entire system by relaxing the data-transfer bottleneck of logic and memory, the proposed MCU configuration achieves ultra-low power consumption and high-speed operation. As confirmed by the results obtained via measurements performed on a fabricated chip, the proposed MCU design, on average, consumed 47.14 $\mu \text{W}$ power at an operating frequency of 200 MHz. This corresponds to the world’s highest signal-processing performance and energy efficiency of highly functional IoT sensor nodes powered by harvested energy

Using several material systems with various magnitudes of the interfacial Dzyaloshinskii-Moriya interaction (DMI), we elucidate a critical influence of domain wall (DW) anisotropy on the current-induced hysteresis loop shift scheme widely employed to determine the magnitude of the Dzyaloshinskii-Moriya effective field (${H}_{\mathrm{DMI}}$). Taking into account the DW anisotropy in the analysis of the hysteresis loop shift, which has not been included in the original model [Phys. Rev. B 93, 144409 (2016)], we show that it provides quantitative agreement of ${H}_{\mathrm{DMI}}$ with that determined from an asymmetric bubble expansion technique for small DMI material systems. For large DMI systems, the DW anisotropy gives rise to nonlinearity in the response of spin-orbit torque efficiency to the in-plane magnetic field, from which ${H}_{\mathrm{DMI}}$ can be determined. The consequence of the directions of DW motion in the Hall device on the current-induced shift of the hysteresis loop is also discussed. The present findings deliver important insights for reliable evaluation of DMI, which are of significance in spintronics with chiral objects.

The mutual synchronization of spin-torque oscillators (STOs) is critical for communication, energy harvesting and neuromorphic applications. Short range magnetic coupling-based synchronization has spatial restrictions (few µm), whereas the long-range electrical synchronization using vortex STOs has limited frequency responses in hundreds MHz (, Spin torque oscillators (STOs) are attractive potential alternative for many high frequency applications, due to their small area and CMOS compatibility. Here, Sharma et al succeed in the electrical synchronization of four STOs, and use their setup to demonstrate wireless and battery-free energy harvesting using eight STOs.

Two-dimensional (2D) van der Waals (vdW) magnetic materials have attracted a lot of attention owing to the stabilization of long range magnetic order down to atomic dimensions, and the prospect of novel spintronic devices with unique functionalities. The clarification of the magnetoresistive properties and its correlation to the underlying magnetic configurations is essential for 2D vdW-based spintronic devices. Here, the effect of Co-doping on the magnetic and magnetotransport properties of Fe3GeTe2 have been investigated. Magnetotransport measurements reveal an unusual Hall effect behavior whose strength was considerably modified by Co-doping and attributed to arise from the underlying complicated spin textures. The present results provide a clue to tailoring of the underlying interactions necessary for the realization of a variety of unconventional spin textures for 2D vdW FM-based spintronics.

We study the timescale of random telegraph noise (RTN) of nanomagnets in stochastic magnetic tunnel junctions (MTJs). From analytical and numerical calculations based on the Landau-Lifshitz-Gilbert and the Fokker-Planck equations, we reveal mechanisms governing the relaxation time of perpendicular easy-axis MTJs (p-MTJs) and in-plane easy-axis MTJs (i-MTJs), showing that i-MTJs can be made to have faster RTN. Superparamagnetic i-MTJs with small in-plane anisotropy and sizable perpendicular effective anisotropy show relaxation times down to 8 ns at negligible bias current, which is more than 5 orders of magnitude shorter than that of typical stochastic p-MTJs and about 100 times faster than the shortest time of i-MTJs reported so far. The findings give a new insight and foundation in developing stochastic MTJs for high-performance probabilistic computers.

Reducing the relaxation time between binary spin states is crucial to enhance the speed and accuracy of a spintronics-based approach for a probabilistic computer, as envisioned by R.P. Feynman. Contrary to previous perception that the reduction of the energy barrier between the states is the most efficient way, the authors discover here that the relaxation time in in-plane easy-axis nanomagnets is widely controllable by changing the effective perpendicular anisotropy field, even though it has nothing to do with the energy barrier.

INTRODUCTION: About one-third of patients with severe ulcerative colitis (UC) do not respond to corticosteroid therapy and receive rescue therapy with infliximab or cyclosporine. Up to 20% of such patients fail to respond to rescue therapy and undergo colectomy. OBJECTIVE: We investigated the outcomes of infliximab and a plant-based diet (PBD) as first-line therapy for severe UC. METHODS: Patients with severe UC defined by the Truelove and Witts criteria were admitted and given standard induction therapy with infliximab (5.0 mg/kg-7.5 mg/kg) at 0, 2, and 6 weeks. Additionally, they received a PBD. The primary endpoint was remission or colectomy in the induction phase and 1 year after discharge. Secondary endpoints were changes in inflammatory markers in the induction phase and the PBD score at baseline and follow-up. A higher PBD score indicates greater adherence to a PBD. RESULTS: Infliximab and PBD as first-line therapy was administered in 17 cases. The remission rate was 76% (13/17), and the colectomy rate was 6% (1/17) in the induction phase. C-reactive protein values and the erythrocyte sedimentation rate significantly decreased at week 6 from 9.42 mg/dL to 0.33 mg/dL and from 59 to 17 mm/h, respectively (p < 0.0001). At 1-year follow-up, the cumulative relapse rate was 25%, and there were no additional colectomy cases. Mean PBD scores of 27.7 at 1 year and 23.8 at 4 years were significantly higher than baseline scores of 8.3 and 9.9, respectively (p < 0.0001 and p = 0.0391). CONCLUSION: This new first-line therapy for severe UC demonstrated a higher remission rate and lower colectomy rate than with the current modality.

We perform hysteresis-loop measurement and domain imaging for [Formula: see text]-oriented D019-Mn3+ xSn1- x ([Formula: see text]) thin films using the magneto-optical Kerr effect (MOKE) and compare it with the anomalous Hall effect (AHE) measurement. We obtain a large Kerr rotation angle of 10 mdeg, comparable with bulk single-crystal Mn3Sn. The composition x dependence of AHE and MOKE shows a similar trend, suggesting the same origin, i.e., the non-vanishing Berry curvature in the momentum space. Magnetic domain observation at the saturated state shows that x dependence of AHE and MOKE is explained by the amount of the reversible area that crucially depends on the crystalline structure of the film. Furthermore, in-depth observation of the reversal process reveals that the reversal starts with nucleation of sub-micrometer-scale domains dispersed in the film, followed by domain expansion, where the domain wall preferentially propagates along the [Formula: see text] direction. Our study provides a basic understanding of the spatial evolution of the reversal of the chiral-spin structure in non-collinear antiferromagnetic thin films.

We investigate the mechanism of analoglike switching of ${\mathrm{Pt}}_{38}{\mathrm{Mn}}_{62}$/[$\mathrm{Co}$/$\mathrm{Ni}$] multilayers induced by spin-orbit torques. X-ray photoemission microscopy performed during magnetization reversal driven by current pulses shows that sequential switching of reproducible domain patterns can be achieved. Switching proceeds by domain-wall displacement starting from the edges of blocked ferromagnetic domains, which do not switch for either direction of the current and represent up to 24% of the total ferromagnetic area. The antiferromagnetic ${\mathrm{Pt}}_{38}{\mathrm{Mn}}_{62}$ layer has a granular texture, with the majority of the domains being smaller than 100 nm, whereas the ferromagnetic domains in $\mathrm{Co}$/$\mathrm{Ni}$ are typically larger than 200 nm. The blocked domains and the granular distribution of exchange bias constrain the origin as well as the displacement of the domain walls, thus leading to highly reproducible switching patterns as a function of the applied current pulses. These measurements clarify the origin of the memristive behavior in antiferromagnet-ferromagnet structures and provide clues for further optimization of spin-orbit torque switching and memristivity in these systems.

Synchronization of large spin Hall nano-oscillator (SHNO) arrays is an appealing approach toward ultrafast non-conventional computing. However, interfacing to the array, tuning its individual oscillators and providing built-in memory units remain substantial challenges. Here, we address these challenges using memristive gating of W/CoFeB/MgO/AlOx-based SHNOs. In its high resistance state, the memristor modulates the perpendicular magnetic anisotropy at the CoFeB/MgO interface by the applied electric field. In its low resistance state the memristor adds or subtracts current to the SHNO drive. Both electric field and current control affect the SHNO auto-oscillation mode and frequency, allowing us to reversibly turn on/off mutual synchronization in chains of four SHNOs. We also demonstrate that two individually controlled memristors can be used to tune a four-SHNO chain into differently synchronized states. Memristor gating is therefore an efficient approach to input, tune and store the state of SHNO arrays for non-conventional computing models.

Electrical manipulation of magnetic materials by current-induced spin torque constitutes the basis of spintronics. Here, we show an unconventional response to spin-orbit torque of a non-collinear antiferromagnet Mn

New methods to induce magnetization switching in a thin ferromagnetic material using femtosecond laser pulses without the assistance of an applied external magnetic field have recently attracted a lot of interest. It has been shown that by optically triggering the reversal of the magnetization in a GdFeCo layer, the magnetization of a nearby ferromagnetic thin film can also be reversed via spin currents originating in the GdFeCo layer. Here, using a similar structure, it is shown that the magnetization reversal of the GdFeCo is not required in order to reverse the magnetization of the ferromagnetic thin film. This switching is attributed to the ultrafast spin current and can be generated by the GdFeCo demagnetization. A larger energy efficiency of the ferromagnetic layer single pulse switching is obtained for a GdFeCo with a larger Gd concentration. Those ultrafast and energy efficient switchings observed in such spintronic devices open a new path toward ultrafast and energy efficient magnetic memories., In GdFeCo/Cu/[Co/Pt] spin valve, the ultrafast demagnetization of the ferrimagnetic GdFeCo alloy, generated by a single femtosecond laser pulse, is shown to generate enough spin current to switch the magnetization of the ferromagnetic Co/Pt multilayer. By increasing the Gd concentration, the Co/Pt switching is found to be more energy efficient.

Arrays of interacting 2D nanomagnets display unprecedented electromagnetic properties via collective effects, demonstrated in artificial spin ices and magnonic crystals. Progress toward 3D magnetic metamaterials is hampered by two challenges: fabricating 3D structures near intrinsic magnetic length scales (sub-100 nm) and visualizing their magnetic configurations. Here, we fabricate and measure nanoscale magnetic gyroids, periodic chiral networks comprising nanowire-like struts forming three-connected vertices. Via block copolymer templating, we produce Ni75Fe25 single-gyroid and double-gyroid (an inversion pair of single-gyroids) nanostructures with a 42 nm unit cell and 11 nm diameter struts, comparable to the exchange length in Ni–Fe. We visualize their magnetization distributions via off-axis electron holography with nanometer spatial resolution and interpret the patterns using finite-element micromagnetic simulations. Our results suggest an intricate, frustrated remanent state which is ferromagnetic but without a unique equilibrium configuration, opening new possibilities for collective phenomena in magnetism, including 3D magnonic crystals and unconventional computing.

We investigated the spin-orbit (SO) effective magnetic-field-induced spin precession of a high-mobility two-dimensional electron gas (2DEG) in a modulation-doped (001) GaAs/AlGaAs quantum well close to the persistent spin helix (PSH) state. The oscillating spin signal induced by the SO fields in zero external magnetic fields was clearly observed. When the photoexcited carrier density is sufficiently small compared with the 2DEG density, the spin precession length in the space domain was obtained by analyzing the precession frequency in the PSH regime. The excitation density dependence of the spin dynamics was reproduced well using the scattering time-dependent spin dynamics simulation. We revealed that the increase in the photoexcited carriers results in a transition from ballistic motion to diffusive motion, leading to a decrease in the spin-diffusion coefficient.

Two-dimensional (2D) van der Waals (vdW) magnetic materials have attracted a lot of attention owing to the stabilization of long-range magnetic order down to atomic dimensions, and the prospect of novel spintronic devices with unique functionalities. The clarification of the magnetoresistive properties and its correlation to the underlying magnetic configurations is essential for 2D vdW-based spintronic devices. Here, the effect of Co-doping on the magnetic and magnetotransport properties of Fe3GeTe2 have been investigated. Magnetotransport measurements reveal an unusual Hall effect behavior whose strength was considerably modified by Co-doping and attributed to arise from the underlying complicated spin textures. The present results provide a clue to tailoring of the underlying interactions necessary for the realization of a variety of unconventional spin textures for 2D vdW FM-based spintronics., Comment: 20 pages, 6 figures

We investigate the physical mechanism governing the sigmoid-like time-averaged response of stochastic magnetic tunnel junctions (s-MTJ), which is a promising building block for probabilistic computers. We measure the time-averaged resistance of perpendicular easy-axis s-MTJs with various free-layer thicknesses and diameters as functions of an external magnetic field and current. The time-averaged response shows no significant dependence on the free-layer thickness, whereas significantly varies with the diameter. Based on the Neel-Arrhenius law, we derive an analytical expression of the time-averaged response against both the magnetic field and current and discuss the underlying mechanism accounting for the obtained results. We show that the experimental results are well explained by considering magnetically active and electrically active volumes of the superparamagnetic free layer in s-MTJs. The obtained finding provides an important design guideline of s-MTJs for probabilistic computers.

Magnetization processes and magnetic domain structures in Ta/CoFeB/MgO stacks were studied in a series of samples with various CoFeB thicknesses d ranging from 1.24 to 1.60 nm with a step of 0.04 nm, using polar magneto-optical Kerr effect (PMOKE) magnetometry and microscopy. Thickness dependence of the magnetic anisotropy was evaluated and the first and second order anisotropy constants were quantified for each thickness. Accordingly, this dependence was deduced to result in magnetization reorientation from out-of-plane to in-plane through an easy-cone magnetization region (1.39 nm ≤ d ≤ 1.41 nm) as d was increased. PMOKE imaging of the magnetization reversal processes for stacks with out-of-plane easy axis indicated both a significant increase of the density of nucleation centers and a change in domain morphology with increasing d up to the magnetization reorientation thickness. Magnetization reversal dynamics was described by a thermal activation model consistent with a Barkhausen length of about 120 nm. The thinnest films with d = 1.24 and 1.28 nm exhibited straightened narrow stripe domains resulting from magnetic dipolar repulsion. A thorough study of narrow stripe domains was performed via direct and indirect magnetization reversal processes. The application of such structures as spin wave nano-channels could be promising.

We investigated an effect of sputtering gas species (Ar, Kr, and Xe) for deposition of a W insertion layer in the CoFeB/W/CoFeB free layer on magnetic properties of the free layer and tunnel magnetoresistance (TMR) ratio of magnetic tunnel junctions (MTJs) stacks using the free layer annealed at 400 °C for 1 h. As the W insertion layer thickness $t_{W}$ increased, we found the degradation of perpendicular anisotropy and larger reduction of saturation magnetic moment per unit area $m_{S}$ in the free layer using Ar compared to those using Kr and Xe. We also found a smaller TMR ratio for the MTJ stack using Ar compared to those using Kr and Xe. Energy-dispersive X-ray spectrometry line analysis revealed more significant interdiffusion between W and CoFeB layers in the free layer using Ar than those using Kr and Xe, that could result in the smaller $m_{S}$ and perpendicular anisotropy in the free layer and smaller TMR ratio for the MTJ stack using Ar than those using Kr and Xe. We also investigated concentration of Ar, Kr, and Xe in W layers deposited using Ar, Kr, and Xe, respectively, by high-resolution Rutherford backscattering spectrometry, revealing that 0.2 at% Ar was detected in the W layer using Ar, while Kr and Xe were not detected in W layers using Kr and Xe. Such a difference in concentration of inert gas atoms in the W layer could be one possible reason for the difference about degree of interdiffusion between W and CoFeB layers.

This article reviews spintronics based memories, in particular, magnetic random access memory (MRAM) in a systematic manner. Debuted as a humble 4 Mb product by FreeScale in 2006, the MRAM has grown to a 256 Mb product of Everspin in 2016. During this period, MRAM has overcome several hurdles and have reached a stage, where the potential for MRAM is very promising. One of the main hurdles that the MRAM overcome between 2006 and 2016 is the way the information is written. The 4 Mb MRAM used a magnetic field based switching technology that would be almost impossible to scale below 100 nm. The 256 Mb MRAM, on the other hand uses a different writing mechanism based on Spin Transfer Torque (STT), which is scalable to very low dimensions. In addition to the difference in the writing mechanism, there has also been a major shift in the storage material. Whereas the 4 Mb MRAM used materials with in-plane magnetic anisotropy, the 256 Mb MRAM uses materials with a perpendicular magnetic anisotropy (PMA). MRAM based on PMA is also scalable to much higher densities. The paper starts with a brief history of memory technologies, followed by a brief description of the working principles of MRAM for novice. Reading information from MRAM, the technologies, materials and the physics behind reading of bits in MRAM are described in detail. As a next step, the physics and technologies involved in writing information are described. The magnetic field based writing and its limitations are described first, followed by an explanation of STT mechanism. The materials and physics behind storage of information is described next. MRAMs with in-plane magnetization, their layered material structure and the disadvantages are described first, followed by the advantages of MRAMs with perpendicular magnetization, their advantages etc. The technologies to improve writability and potential challenges and reliability issues are discussed next. Some of the future technologies that might help the industry to move beyond the conventional MRAM technology are discussed at the end of the paper, followed by a summary and an outlook. Nanyang Technological University Published version The authors (SNP and SB) acknowledge the funding support of the start-up grant of Nanyang Technological University, Singapore.

We investigate the stack structure dependence of magnetic properties on thin films that consist of an antiferromagnetic PtMn and a ferromagnetic Co/Ni multilayer for field-free spin-orbit torque-induced magnetization switching devices. Magnetic parameters, such as the spontaneous magnetization, effective and interfacial magnetic anisotropies, and exchange bias field are quantified as a function of stack structure. Engineering of the stack allows the improvement of current-induced magnetization switching characteristics compared with a previous work, which is confirmed using patterned Hall cross devices.

Studies of magnetization dynamics have incessantly facilitated the discovery of fundamentally novel physical phenomena, making steady headway in the development of magnetic and spintronics devices. The dynamics can be induced and detected electrically, offering new functionalities in advanced electronics at the nanoscale. However, its scattering mechanism is still disputed. Understanding the mechanism in thin films is especially important, because most spintronics devices are made from stacks of multilayers with nanometer thickness. The stacks are known to possess interfacial magnetic anisotropy, a central property for applications, whose influence on the dynamics remains unknown. Here, we investigate the impact of interfacial anisotropy by adopting CoFeB/MgO as a model system. Through systematic and complementary measurements of ferromagnetic resonance (FMR), on a series of thin films, we identify narrower FMR linewidths at higher temperatures. We explicitly rule out the temperature dependence of intrinsic damping as a possible cause, and it is also not expected from existing extrinsic scattering mechanisms for ferromagnets. We ascribe this observation to motional narrowing, an old concept so far neglected in the analyses of FMR spectra. The effect is confirmed to originate from interfacial anisotropy, impacting the practical technology of spin-based nanodevices up to room temperature., 23 pages,3 figures

The importance of ferromagnetic resonance, discussed more than half a century ago by Charles Kittel, is ever-increasing, as we use it to evaluate various forms of spin-orbit torque in spintronics. However, a widely employed method can lead to significant overestimation of the torque in some systems. The authors offer a scheme that allows the quantification of spin-orbit torques, free from spurious signals---a welcome contribution to the worldwide effort to create the next generation of ultralow-power devices for data storage and processing.

The quantum Hall effect (QHE) is example of dissipationless electrical transport that can occur when a large magnetic field is applied to a high mobility two-dimensional electron gas at absolute zero. As QHE depends on a large applied magnetic field, it was natural to ask if a material could possess similar quantized transport inherently, without Landau levels from external fields. The surface states under the magnetic domains can be gapped by exchange coupling, and acquire a mass, whose sign is dependent on the direction of magnetization of the overlying domain. One way to accomplish this is to make the TI itself ferromagnetic by magnetic doping, and to control the carrier density so that chemical potential lies inside the magnetic gap. The choice of magnetic dopant is thus crucial for the realization of the quantum anomalous Hall effect. Ferromagnetism arises in dilute magnetic semiconductors via the carrier-mediated Ruderman–Kittel–Kasuya–Yosida exchange interaction.

CONTEXT: No known previous study has focused on plant-based diet (PBD) to prevent relapse of ulcerative colitis (UC) except our previous educational hospitalization study. OBJECTIVE: To describe the relapse rate in a large case series of UC after incorporation of PBD into induction therapy. DESIGN: All patients with UC between 2003 and 2017 were admitted for induction therapy. Patients receiving educational hospitalization or treated with infliximab were excluded. A lacto-ovo-semivegetarian diet (PBD) together with medication prescribed according to UC guidelines was provided during hospitalization. MAIN OUTCOME MEASURES: The primary endpoint was relapse during follow-up. The secondary endpoint was change over time in the plant-based diet score (PBDS), which evaluated adherence to the PBD. RESULTS: Ninety-two cases were studied, of which 51 were initial episodes and 41 were relapses. Cases varied in severity (31 mild, 48 moderate, 13 severe) and extent (15 proctitis, 22 left-sided colitis, 55 extensive colitis). More severe cases existed among the relapse cases than among the initial episode cases. Cumulative relapse rates at 1- and 5-year follow-up (Kaplan-Meier analysis) were 14% and 27%, respectively, for the initial episode cases, and 36% and 53%, respectively, for relapse cases. At long-term follow-up (6 years 4 months), PBDS was significantly higher than baseline PBDS (p < 0.0001). CONCLUSION: Relapse rates in UC after induction therapy with PBD were far lower than those previously reported with conventional therapy. Adherence to PBD was significantly higher than baseline even at 6-year follow-up. We conclude PBD is effective for preventing UC relapse. (Study identification no.: UMIN000019061: Registration: www.umin.ac.jp)

Recently, the demand for low-power, high-performance microcontroller units (MCUs) for power-supply-critical sensor node applications has been increasing. In response to this demand, the use of nonvolatile memory elements for realizing MCUs for sensor node applications has been actively researched and developed. The latest nonvolatile MCUs (NV-MCUs) demonstrated 32b operation at 30MHz [1] and 8b operation at 100MHz [2]. However, this performance level is not suitable for sensor node applications that process large numbers of received signals and extract valuable information from them immediately to reduce the amount of transfer data to a data center. The use of various nonvolatile devices has also been proposed. However, these devices exhibit critical drawbacks when applied to sensor node applications, including limited endurance and low compatibility with standard CMOS. A spintronics-based nonvolatile device with unlimited endurance, a short switching time, and CMOS compatibility is a promising candidate for designing a low-power, high-performance NV-MCU.

Conventional computers operate deterministically using strings of zeros and ones called bits to represent information in binary code. Despite the evolution of conventional computers into sophisticated machines, there are many classes of problems that they cannot efficiently address, including inference, invertible logic, sampling and optimization, leading to considerable interest in alternative computing schemes. Quantum computing, which uses qubits to represent a superposition of 0 and 1, is expected to perform these tasks efficiently1–3. However, decoherence and the current requirement for cryogenic operation4, as well as the limited many-body interactions that can be implemented, pose considerable challenges. Probabilistic computing1,5–7 is another unconventional computation scheme that shares similar concepts with quantum computing but is not limited by the above challenges. The key role is played by a probabilistic bit (a p-bit)—a robust, classical entity fluctuating in time between 0 and 1, which interacts with other p-bits in the same system using principles inspired by neural networks8. Here we present a proof-of-concept experiment for probabilistic computing using spintronics technology, and demonstrate integer factorization, an illustrative example of the optimization class of problems addressed by adiabatic9 and gated2 quantum computing. Nanoscale magnetic tunnel junctions showing stochastic behaviour are developed by modifying market-ready magnetoresistive random-access memory technology10,11 and are used to implement three-terminal p-bits that operate at room temperature. The p-bits are electrically connected to form a functional asynchronous network, to which a modified adiabatic quantum computing algorithm that implements three- and four-body interactions is applied. Factorization of integers up to 945 is demonstrated with this rudimentary asynchronous probabilistic computer using eight correlated p-bits, and the results show good agreement with theoretical predictions, thus providing a potentially scalable hardware approach to the difficult problems of optimization and sampling. A probabilistic computer utilizing probabilistic bits, or p-bits, is implemented with stochastic nanomagnetic devices in a neural-network-inspired electrical circuit operating at room temperature and demonstrates integer factorization up to 945.

We investigate the magnetization switching via a combination of spin-transfer torque (STT) and spin–orbit torque (SOT). STT and SOT are simultaneously induced by a pulsed current flowing through an in-plane easy-axis magnetic tunnel junction and an underneath Ta/W channel. SOT allows the magnetization to be switched with the sub-ns pulse down to 200 ps and STT eliminates the necessity of an external field. The switching current is much smaller than the case driven solely by STT in the short pulse regime. We also compare the threshold current between two structures having orthogonal (Type Y) and collinear (Type X) magnetic easy axes to the longitudinal direction of the channel and find that the Type X achieves smaller switching current by a factor of 1/4 at 200 ps.

A shape-anisotropy magnetic tunnel junction (MTJ) holds promise for its scaling into single-digit nanometers while possessing high data-retention capability. Understanding magnetization reversal mode is crucial to quantify the thermal stability factor Δ for data retention with high accuracy. Here, we study magnetization reversal mode in the shape-anisotropy MTJ with a 15-nm-thick CoFeB layer by evaluating Δ from two different methods: switching probability and retention time measurements. We find that magnetization reversal coherently proceeds in the 15-nm-thick and X/1X-nm-diameter cylindrical nanomagnet in the shape-anisotropy MTJs, in contrast to the conventional interfacial-anisotropy MTJs with a smaller thickness and larger diameter. The coherent magnetization reversal of the shape-anisotropy MTJ is also confirmed by astroid curve measurements. This study provides insight into the development of ultrasmall and high-reliability MTJ devices.

Shape-anisotropy magnetic tunnel junctions (MTJs) are attracting much attention as a high-performance nonvolatile spintronic device in the X/1X nm regime. In this study, we investigate an energy barrier relevant to the retention property in CoFeB/MgO-based shape-anisotropy MTJs with various diameters at high temperatures and compare it with that in conventional interfacial-anisotropy MTJs. We find that the scaling relationship between the energy barrier and the spontaneous magnetization in shape-anisotropy MTJs is well described by a model assuming the dominant contribution of shape anisotropy to the energy barrier. Also, the scaling exponent is much smaller than that for the interfacial-anisotropy MTJs, indicating that the properties of shape-anisotropy MTJs are less sensitive to the temperature. Using the experimentally determined scaling relationship, we discuss the design window of the MTJ dimensions to achieve data retention of 10 years at various temperatures. This study demonstrates that the shape-anisotropy MTJ holds promise of scaling beyond 20 nm for high-temperature applications.

Nonvolatile spintronic devices have potential advantages, such as fast read/write and high endurance together with back-end-of-the-line compatibility, which offers the possibility of constructing not only stand-alone RAMs and embedded RAMs that can be used in conventional VLSI circuits and systems but also standby-power-free high-performance nonvolatile CMOS logic employing logic-in-memory architecture. The advantages of employing spintronic devices, especially magnetic tunnel junction (MTJ) devices with CMOS circuits, are discussed, and the current status of the MTJ-based VLSI computing paradigm is presented along with its prospects and remaining challenges.

We investigate the effect of the buffer-layer materials and their crystallographic structures on the magnetic properties of the CoFeB–MgO stacks. Amorphous Ta, amorphous Mo, and crystalline Mo are used as the buffer layer of CoFeB–MgO. The CoFeB–MgO stacks on both the Mo buffer layers show higher perpendicular anisotropy than that on Ta after 400-°C annealing. Difference is also seen between the samples with amorphous and crystalline Mo in the intermixing property between Mo and CoFeB; intermixing is much more unlikely for the crystalline sample. Magnetic damping constant is also evaluated from ferromagnetic resonance measurements. Samples with crystalline Mo buffer layer show smaller damping constant than those with Ta buffer layer after 400-°C annealing. This paper clarifies that the magnetic properties of CoFeB–MgO depend on the buffer-layer materials and their crystallinities.

We investigated annealing temperature $T_{a}$ dependence of tunnel magnetoresistance (TMR) ratio and magnetic properties for perpendicular-anisotropy (CoFe)100– X B X /MgO magnetic tunnel junctions (MTJs) with single (CoFe)100– X B X /MgO interface (s-MTJ) and double CoFeB–MgO interface (d-MTJ) structures with various boron compositions X. High TMR ratio over 100% was observed in the s-MTJ with $X= 35$ at.% after annealing at 360 °C–400 °C, whereas the s-MTJ with $X = 30$ at.% showed the degradation of TMR ratio with the increase of $T_{a}$ above 360 °C, resulting from the decrease of perpendicular anisotropy. The d-MTJ with $X = 25$ at.% maintained high TMR ratio up to $T_{a} = 400~^{\circ }\text{C}$ owing to its higher perpendicular anisotropy compared with the s-MTJ. The difference of perpendicular anisotropy between the s-MTJ and the d-MTJ can be attributed to higher interfacial anisotropy together with lower saturation magnetization of the d-MTJs. The lower saturation magnetization is attributable to two MgO layers that suppress boron diffusion from CoFeB layers, which was verified by cross-sectional line analysis using electron energy-loss spectroscopy.

We investigate magnetic properties of CoFeB/Ta/[Co/Pd (Pt)] multilayers and properties of magnetic tunnel junctions (MTJs) with the structures as a recording layer. CoFeB/Ta/[Co/Pd] multilayer shows a lower damping constant $\alpha $ than that of CoFeB/Ta/[Co/Pt] multilayer. We evaluate current-induced magnetization switching (CIMS) properties of the MTJs with CoFeB/Ta/[Co/Pd (Pt)] multilayers with the junction diameter of 15 (13) nm that show similar thermal stability factor. CIMS is observed for the MTJ with a CoFeB/Ta/[Co/Pd] multilayer at zero magnetic field, whereas it is observed for the MTJ with the CoFeB/Ta/[Co/Pt] multilayer only in the presence of an external field.