R&D: Five Articles on Memories Technologies and Research

R&D: Diamond-based Non-volatile Memory
Findings establish diamond as a viable material for ultra-stable memory applications and provide key insights into its long-term performance.

Carbon has published an article written by Martin Kah, Univ. Grenoble Alpes, CNRS, Grenoble INP, Institut Néel, 38000 Grenoble, France, LAPLACE, Université de Toulouse, CNRS, 31000 Toulouse, France, and Cambridge University, Department of Engineering, Electrical Engineering Division, Cambridge, UK, Cédric Masante Univ. Grenoble Alpes, CNRS, Grenoble INP, Institut Néel, 38000 Grenoble, France, and LAPLACE, Université de Toulouse, CNRS, 31000 Toulouse, France, Nicolas Rouger, LAPLACE, Université de Toulouse, CNRS, 31000 Toulouse, France, Fabrice Donatini, Univ. Grenoble Alpes, CNRS, Grenoble INP, Institut Néel, 38000 Grenoble, France, Juliette Letellier,DIAMFAB, 25 avenue des Martyrs, 38000 Grenoble, France, Franz A Koeck, Robert J Nemanich, Arizona State University, Tempe, AZ 85281, USA, and Julien Pernot, Univ. Grenoble Alpes, CNRS, Grenoble INP, Institut Néel, 38000 Grenoble, France, and Institut Universitaire de France (IUF), 75231 Paris, France.

Abstract: “The digital age has brought unprecedented challenges in data storage, with exponential growth in data generation outpacing traditional storage solutions in capacity, speed, and reliability. Long-term data preservation is particularly concerning, as current technologies like semiconductor drives and hard disk drives struggle with data longevity beyond a few decades. Optoelectronic memories show promise among emerging storage-class memory technologies, but they face trade-offs between information retention time and fatigue cycles due to material limitations. Diamond, with its exceptional physical properties, especially its ultra-wide bandgap, emerges as a strong candidate for developing non-volatile memories capable of retaining information over extremely long periods. This paper explores the fundamental aspects and fabrication methods of diamond non-volatile memories, focusing on the information storage mechanisms and unique properties of the ultra-deep nitrogen donor in diamond, as well as the associated junction field effect transistor. The study describes the physical concept of the memory effect and the transistor structure, details the fabrication and properties of the samples, and addresses key challenges in creating efficient devices. These findings establish diamond as a viable material for ultra-stable memory applications and provide key insights into its long-term performance.“

R&D: Accelerating Erasure Coding on Persistent Memory via Adaptive Prefetcher Scheduling
Authors propose Dialga, an adaptive hardware/software prefetching scheduler aware of PM encoding characteristics.

ACM Digital Library has published, in ICPP ’25: Proceedings of the 54th International Conference on Parallel Processing, an article written by Guanglei Xu, Hai Zhou, Yuchong Hu, Dan Feng, and Renzhi Xiao,Huazhong University of Science and Technology, Wuhan, China.

Abstract: “Compared to DRAM, persistent memory (PM) offers higher density and persistence but encounters more severe reliability challenges. Erasure coding is widely adopted to enhance reliability with minimal space overhead. Unfortunately, applying erasure coding to PM introduces significant additional latency. Previous work to mitigate coding latency has primarily focused on optimizing computational efficiency. Instead, we reveal that the main performance bottleneck is high memory latency due to inefficient hardware prefetchers, rather than computation. We further observe that the prefetching inefficiency mainly results from: (i) too wide or narrow coding stripes, (ii) small block sizes, and (iii) high concurrency.“

“Based on these observations, we propose Dialga, an adaptive hardware/software prefetching scheduler aware of PM encoding characteristics. Dialga improves memory access efficiency and thereby enhances encoding performance. It first profiles the cache status and access patterns to adaptively switch prefetching strategies. Dialga then employs a lightweight operator to achieve fine-grained and low-overhead scheduling for both hardware and software prefetchers. Additionally, Dialga further optimizes the PM read buffer efficiency by leveraging the implicit data loading mechanism for prefetching. Compared with state-of-the-art erasure coding libraries, Dialga achieves up to 96.6% higher encoding throughput and up to 178.8% improvement in multi-thread scalability.“

R&D: Unveiling Laser-induced Ultrafast Switching Mechanism in Ferromagnetic Spin Valves
Work establishes a robust theoretical framework for optically induced ultrafast magnetization reversal and provides critical insights for the design of future picosecond-scale, low-power, and nonvolatile magnetic recording devices.

Applied Physics Letters has published an article written by Shen Li, Suteng Zhao, State Key Laboratory of Spintronics, Hangzhou International Innovation Institute, Beihang University, Hangzhou 311115, China, and Fert Beijing Institute, MIIT Key Laboratory of Spintronics, School of Integrated Circuit Science and Engineering, Beihang University, Beijing 100191, China, Kunlong Pan, State Key Laboratory of Spintronics, Hangzhou International Innovation Institute, Beihang University, Hangzhou 311115, China, Chen Lv, Fert Beijing Institute, MIIT Key Laboratory of Spintronics, School of Integrated Circuit Science and Engineering, Beihang University, Beijing 100191, China, Wei Yang, State Key Laboratory of Spintronics, Hangzhou International Innovation Institute, Beihang University, Hangzhou 311115, China, and Fert Beijing Institute, MIIT Key Laboratory of Spintronics, School of Integrated Circuit Science and Engineering, Beihang University, Beijing 100191, China, Pierre Vallobra, State Key Laboratory of Spintronics, Hangzhou International Innovation Institute, Beihang University, Hangzhou 311115, China, Wei Zhang, State Key Laboratory of Spintronics, Hangzhou International Innovation Institute, Beihang University, Hangzhou 311115, China, and Fert Beijing Institute, MIIT Key Laboratory of Spintronics, School of Integrated Circuit Science and Engineering, Beihang University, Beijing 100191, China, Luding Wang, Center for Science and Innovation in Spintronics, Tohoku University, Sendai 980-8577, Japan, Konstantin A. Zvezdin, “New Spintronic Technologies” Limited Liability Company, 121205 Skolkovo, Moscow, Russia, Xiaoyang Lin, and Weisheng Zhao, State Key Laboratory of Spintronics, Hangzhou International Innovation Institute, Beihang University, Hangzhou 311115, China, and Fert Beijing Institute, MIIT Key Laboratory of Spintronics, School of Integrated Circuit Science and Engineering, Beihang University, Beijing 100191, China.

Abstract: “The recent demonstration of single-shot ultrafast magnetization reversal in ferromagnetic spin valves—combining spin-transfer torque with optically induced ultrafast switching—has offered a promising avenue for next-generation magnetic storage technologies. However, a comprehensive theoretical framework is currently lacking to validate and elucidate the reversal mechanisms across different initial magnetic states. Here, we develop a theoretical model for optically induced ultrafast magnetization reversal by integrating the s-d exchange model with an atomistic spin dynamics approach. The proposed model’s validity is corroborated through detailed comparisons with experimental time-resolved magneto-optic Kerr effect data. Our findings highlight distinct contributions from ultrafast demagnetization and ultrafast spin currents to the switching process. Furthermore, we systematically explore the influence of laser pulse parameters, such as fluence and width, as well as material-specific properties like magnetic anisotropy and Gilbert damping coefficients on ultrafast ferromagnetic reversal. Our findings indicate that increasing laser pulse fluence intensifies ultrafast demagnetization and enhances spin current strength, whereas extending pulse width delays demagnetization and diminishes spin current intensity. Notably, magnetic anisotropy exerts minimal influence on spin current generation, while higher damping coefficients amplify spin current intensity, thereby facilitating ultrafast reversal. Comparative simulations across various spin valve materials reveal that CoFe exhibits superior ultrafast spin current conversion efficiency compared to [Co/Ni]_n and CoPt-based systems. This work establishes a robust theoretical framework for optically induced ultrafast magnetization reversal and provides critical insights for the design of future picosecond-scale, low-power, and nonvolatile magnetic recording devices.“

R&D: Effect of Different Top Electrodes and Magnetic Field on HfO_X based RRAM Devices
Resistance tuning confirms that the magnetic field can improve the RRAM device’s switching process when suitable electrodes are used, and it also paves the way for emerging multilevel non-volatile memory technology.

Journal of Physics D: Applied Physics, Special Issue on Growth of 2D Layers and Electronic Materials – Papers from iCOLD25, has published an article written by Bhaswati Biswas, Abhijit Ghosh, Syed Jalaluddeen, Rohiteswar Mondal, Chandrasekhar Murapaka, and Suresh Kumar Garlapati, Department of Materials Science and Metallurgical Engineering, Indian Institute of Technology Hyderabad, Kandi, 502284 Sangareddy, Telangana, India.

Abstract: “Hafnium oxide (HfO_X) is one of the best metal oxides for resistive random-access memory (RRAM) devices, however different electrode configurations and the effect of magnetic field on these devices are yet to be studied. In this work, we report sputter-deposited HfO_X-based RRAM devices with different top electrodes (TEs) and the combined effects of magnetic and electric fields on the switching behavior. The RRAM devices have been fabricated with a common bottom electrode, indium tin oxide, and different TEs such as Ag, Pt, and Ni. All the prepared devices have demonstrated bipolar resistive switching mode. The active electrode Ag has shown good endurance and retention time of 10² cycles and 10³ s, respectively, and has a wider memory window of >10³. On the other hand, devices with an inert electrode, Pt, have a lower operating voltage. Furthermore, devices with Ni TE exhibit the highest stability over cycle-to-cycle (up to 10³ cycles), compared to the other two TE configurations. Further, we have investigated the effect of the magnetic field on the RS in our devices with Ag and Ni TEs. Interestingly, due to the induced magnetoresistance effect, the high resistance state and low resistance state values increased simultaneously in the presence of magnetic field. For Ni-based devices, the magnetic field effect is more pronounced than in other devices, and the resistance ratio enhances from ∼1.7 to ∼2.6 in the presence of magnetic field. Additionally, the set and reset voltages shifted toward higher values due to the Lorentz force. The resistance tuning confirms that the magnetic field can improve the RRAM device’s switching process when suitable electrodes are used, and it also paves the way for emerging multilevel non-volatile memory technology.“

R&D: Emerging Transistors and Magnetic Memories for Harsh Radiation Environments
Review systematically evaluates recent progress in next-generation radiation-tolerant transistors and magnetic memory technologies, particularly emphasis on carbon nanotube field-effect transistors (CNTFETs), two-dimensional material-based FETs (2D-FETs), spin-transfer torque magnetic random-access memory (STT-MRAM), and spin–orbit torque MRAM (SOT-MRAM).

Journal of Physics: Condensed Matter has published an article written by Daming Zhou, Fert Beijing Institute, School of Integrated Circuit Science and Engineering, Beihang University, Beijing, People’s Republic of China, and State Key Laboratory of Spintronics, Hangzhou International Innovation Institute, Beihang University, Hangzhou, People’s Republic of China, Xinyu Wang, Fert Beijing Institute, School of Integrated Circuit Science and Engineering, Beihang University, Beijing, People’s Republic of China, Ke Zhang, Xiaoyang Lin, and Weisheng Zhao, Fert Beijing Institute, School of Integrated Circuit Science and Engineering, Beihang University, Beijing, People’s Republic of China, and State Key Laboratory of Spintronics, Hangzhou International Innovation Institute, Beihang University, Hangzhou, People’s Republic of China.

Abstract: “The development of radiation-tolerant electronics is indispensable for aerospace, deep-space missions, nuclear power infrastructure, and medical devices, where conventional silicon-based systems suffer severe performance degradation in harsh radiation environments. This review systematically evaluates recent progress in next-generation radiation-tolerant transistors and magnetic memory technologies, particularly emphasis on carbon nanotube field-effect transistors (CNTFETs), two-dimensional material-based FETs (2D-FETs), spin-transfer torque magnetic random-access memory (STT-MRAM), and spin–orbit torque MRAM (SOT-MRAM). We analyze their underlying operating mechanisms, notable breakthroughs, radiation tolerance, and scalability potential relative to silicon FETs (Si-FETs) and volatile memory counterparts such as SRAM. For example, the atomic-scale thin channels and robust covalent bonding in CNTFETs and 2D-FETs inherently suppress radiation-induced displacement damage, while spin-based STT-/SOT-MRAM demonstrates intrinsic radiation resistance through magnetism-mediated data storage. Nevertheless, persistent challenges in material deposition, interface optimization, and manufacturing scalability hinder practical implementation. Additionally, monolithic three-dimensional (M3D) integration as a transformative approach has the potential to assemble these emerging technologies into radiation-hardened systems with superior functional density. By enabling vertical stacking of heterogeneous devices and compact interconnects, M3D architectures could overcome traditional scaling bottlenecks while synergizing the radiation tolerance of carbon nanomaterials, 2D semiconductors, and magnetic memory elements. This review outlines a strategic roadmap for next-generation radiation-tolerant electronics, highlighting critical innovations across materials science, device physics, and advanced integration paradigms.“