Fact-checked by Grok 2 weeks ago

Non-volatile memory

Non-volatile memory (NVM), also referred to as non-volatile storage, is a type of that retains stored data even after the power supply is disconnected, in contrast to like (DRAM) which loses its contents without continuous power. The development of NVM has roots in early mechanical systems from the 19th century, such as punched cards and tape, with semiconductor-based NVM emerging in the 1960s through charge-based devices, evolving through (PROM) and (EPROM) in the 1970s, and reaching widespread adoption with (EEPROM) and in the 1980s. Common types include mechanically addressed systems such as (e.g., hard disk drives) and optical media (e.g., CDs); (ROM) variants such as mask ROM and PROM for fixed data storage; and , which allow electrical rewriting and are foundational to solid-state drives (SSDs) and USB flash drives; and emerging technologies like (FeRAM), magnetic RAM (MRAM), phase-change memory (PCM), and resistive RAM (ReRAM), which offer potential improvements in speed, density, and energy efficiency for future applications. NVM plays a critical role in modern , serving as secondary for operating systems, applications, and user in personal computers, smartphones, and ; enabling boot firmware in devices like /; and supporting systems in automotive, industrial, and for persistent configuration and code . In enterprise environments, NVM technologies such as NAND in SSDs provide high-capacity, durable with faster access times than traditional hard disk drives, driving advancements in data centers and . Ongoing research into emerging NVM aims to bridge the performance gap with while maintaining persistence, potentially revolutionizing fields like and in-memory databases.

Introduction

Definition and characteristics

Non-volatile memory (NVM) is a type of that retains stored data even after the power supply is disconnected, encompassing both electrically addressed technologies and mechanically addressed systems such as magnetic and . This distinguishes it from , which loses data without continuous . Persistence arises from the memory's ability to store data in stable physical states that do not dissipate upon power removal. Key characteristics of NVM include high data persistence, with typical retention times ranging from 10 years to several decades under standard operating conditions, serving as a primary measure of non-volatility through the duration data can be held without power or refresh. Access times generally fall between tens of microseconds and milliseconds, balancing speed with reliability. Storage density spans from individual bits in systems to terabits in high-capacity devices, enabling scalable data handling. , defined as the number of reliable read/write cycles per , varies widely from approximately $10^3 to $10^{15}, depending on the underlying and influencing suitability for frequent updates. Power consumption is negligible during retention (effectively zero, as no energy is needed to preserve states) but occurs primarily during read and write operations, often in the range of microwatts to milliwatts per access. At the fundamental level, NVM operates by leveraging durable physical phenomena to encode states, such as charge in insulating layers (where electrons are confined in potential wells), magnetic orientation in ferromagnetic materials (aligning spins to represent ), phase changes in chalcogenide alloys (switching between amorphous and crystalline forms), or variable resistance states in metal-oxide films (altering without ongoing power). These mechanisms ensure bistable or multistable configurations that remain intact post-power-off, contrasting with volatile systems reliant on dynamic charge refresh. in NVM is organized in basic units: a bit (0 or 1), with 8 bits forming a byte; larger blocks like sectors (typically 512 bytes or multiples thereof) facilitate efficient addressing and error management. For charge-based NVM, is often modeled using an framework to predict loss over time due to leakage currents. A simplified equation for retention time t_{ret} is t_{ret} = \tau \ln \left( \frac{V_{initial}}{V_{threshold}} \right), where \tau represents the characteristic influenced by material properties and , V_{initial} is the programmed voltage level, and V_{threshold} is the minimum voltage distinguishing states. This model highlights how retention degrades exponentially with factors like elevated temperatures or cell wear, guiding reliability assessments.

Comparison to volatile memory

Volatile memory, such as (SRAM) and (DRAM), is characterized by its rapid data access times, typically in the range of 10-50 nanoseconds for DRAM, making it ideal for high-speed caching and temporary data storage in computing systems. However, these technologies require continuous to maintain stored data; upon power loss, all information is erased, as the memory cells rely on active electrical charges or currents to retain . In contrast, non-volatile memory (NVM) provides persistence, retaining data without power, which is a fundamental advantage over volatile memory's on power-off, enabling reliable long-term storage for applications like boot devices and in embedded systems. While volatile memory excels in read/write speeds—often orders of magnitude faster than NVM, with achieving sub-nanosecond access—NVM offers higher storage density and lower standby power consumption, though at the cost of slower access times and limited write endurance. For instance, provides effectively unlimited write cycles and lower cost per compared to many NVM types, whereas NVM like achieves densities up to terabits per chip but with endurance limited to 10,000-100,000 cycles for typical consumer grades. These trade-offs influence their use cases: volatile memory supports transient computations in CPUs and , while NVM serves persistent needs in solid-state drives and . The following table summarizes key trade-offs between representative volatile and non-volatile memory types, highlighting their performance characteristics:
Memory TypeAccess TimeEndurance (Write Cycles)Cost per GigabytePower Consumption (Standby)
(Volatile)10-50 nsEffectively unlimitedLower (~$3-5/GB as of 2025)Higher (requires refresh)
SRAM (Volatile)<1 nsEffectively unlimitedHigher than DRAMModerate
NAND Flash (NVM)10-100 μs (read), 0.1-1 ms (write)10^4-10^5~$0.05-0.1/GB as of 2025Very low
NOR Flash (NVM)50-100 ns (read), 5-10 μs (write)10^5-10^6Higher than NANDLow
Hybrid systems often combine volatile and non-volatile memory in a memory hierarchy, where volatile components handle fast, temporary operations and NVM provides backing storage for data persistence, optimizing overall system performance and efficiency.

History

Early mechanical and pre-semiconductor developments

The earliest forms of non-volatile data storage emerged in the pre-20th century through mechanical systems designed for pattern control and information retention. In 1801, Joseph Marie Jacquard invented the Jacquard loom, which utilized punched cards to automate the weaving of complex textile patterns, marking the first practical use of punched media for storing and retrieving instructional data without power dependency. These cards, made of stiff paper or cardboard with holes representing binary-like instructions, allowed for repeatable, durable data retention that persisted independently of any energy source. Building on this concept, paper tape emerged as another mechanical storage medium; in 1857, Charles Wheatstone adapted perforated paper strips for telegraphy, enabling the preparation, storage, and transmission of sequential data in a compact, non-volatile format suitable for long-term retention. Entering the early 20th century, magnetic recording techniques introduced more dynamic non-volatile storage options, though some hybrid systems retained mechanical elements. In 1898, Danish engineer Valdemar Poulsen developed the telegraphone, the first magnetic wire recorder, which captured audio signals by magnetizing a steel wire with an electromagnet, providing durable, erasable storage that retained data without continuous power. This innovation laid foundational principles for magnetic media, demonstrating non-volatility through hysteresis in ferromagnetic materials. In 1928, German engineer Fritz Pfleumer advanced the field by patenting magnetic tape, consisting of paper strips coated with iron oxide particles, which offered a flexible, low-cost alternative for recording and storing data magnetically. By 1932, Austrian inventor Gustav Tauschek patented the magnetic drum memory, a rotating cylinder coated with ferromagnetic material that stored binary data via magnetic domains, enabling random access in early computing prototypes and serving as a non-volatile auxiliary storage device. In contrast, mercury delay line memory, developed in the 1940s by researchers like J. Presper Eckert for systems such as ENIAC, relied on acoustic waves propagating through liquid mercury tubes to recirculate data; while mechanically implemented, it was inherently volatile, as information decayed without active regeneration, highlighting the limitations of purely acoustic-mechanical approaches. The 1950s saw refinements in magnetic non-volatile storage, bridging mechanical designs toward more efficient electrical integration. In 1949, Chinese-American physicist An Wang invented the core-and-coil pulse transfer device, enabling practical , where tiny ferrite rings stored bits through directional magnetization, offering non-volatility, high reliability, and faster access than prior mechanical systems; this technology was notably employed in computers like the UNIVAC series for main memory applications. Magnetic-core systems, though semi-mechanical in assembly, proved robust for data retention but remained bulky due to the need for extensive wiring and cores. Concurrently, in 1956, IBM introduced the 305 RAMAC (Random Access Method of Accounting and Control), the first commercial , featuring 50 spinning aluminum platters with magnetic surfaces that provided 5 megabytes of non-volatile storage, revolutionizing data handling for business applications despite its room-sized footprint and high cost. These mechanical and early magnetic systems, while pioneering non-volatile data persistence, faced inherent constraints that spurred the shift to semiconductor technologies. Their large physical size—often requiring dedicated rooms—coupled with slow mechanical access speeds (milliseconds for drums and disks) and reliability issues from wear on moving parts like wires, tapes, and rotors, limited scalability for growing computational demands. By the mid-1950s, these drawbacks underscored the need for compact, faster, and more dependable electrical alternatives, setting the stage for semiconductor innovations.

Semiconductor non-volatile memory evolution

The development of semiconductor non-volatile memory (NVM) began in the 1960s, driven by the need for reliable data retention in integrated circuits beyond volatile alternatives. In 1967, Dawon Kahng and Simon Sze at Bell Laboratories invented the floating-gate MOSFET, introducing the first practical non-volatile semiconductor memory cell capable of charge storage for read-only memory (ROM) applications, which laid the foundation for subsequent erasable technologies. This innovation built on Kahng's earlier co-invention of the MOSFET in 1959 with Mohamed Atalla, a device central to modern semiconductors and recognized in the 2000 Nobel Prize in Physics for enabling semiconductor scaling, including NVM advancements. Bipolar programmable read-only memory (PROM) emerged around the same period, allowing one-time programming via fuse links in integrated circuits. The 1970s marked progress in erasable NVM with ultraviolet light-based erasure. In 1971, Dov Frohman at Intel developed the erasable programmable read-only memory (EPROM), utilizing a floating-gate structure to enable UV-erasure and electrical programming, which became a cornerstone for firmware storage. By 1978, Eli Harari at Hughes Aircraft (later Intel) invented electrically erasable PROM (EEPROM), allowing byte-level electrical erasure and reprogramming without external light, addressing EPROM's limitations for flexible data updates. Flash memory revolutionized density and cost in the 1980s. In 1984, Fujio Masuoka at presented NOR flash, a block-erasable architecture using floating gates for high-speed random access, building on earlier project work from 1980. Masuoka's team advanced this to NAND flash in 1987, enabling higher density through serial access and multi-level cells, which propelled mass storage in consumer devices. The 1990s and 2000s saw diversification into alternative NVM types to overcome flash scaling limits like charge leakage. In 1993, Ramtron International commercialized ferroelectric RAM (), leveraging ferroelectric materials for non-volatile, low-power operation in embedded systems. IBM demonstrated the first magnetoresistive RAM () prototype in 1996, using magnetic tunnel junctions for spin-based data retention, targeting cache-like performance. Throughout the 2000s, IBM advanced phase-change memory () research, achieving prototypes with chalcogenide materials for fast switching and endurance exceeding flash. From the 2010s onward, scaling and integration drove commercialization of emerging NVM. In 2008, HP Labs (in collaboration with Intel) demonstrated ReRAM using memristor-based crossbar arrays, enabling dense, low-voltage resistive switching for beyond-3D architectures. Everspin Technologies began production of spin-transfer torque MRAM (STT-MRAM) in 2016, offering 256 Mb densities with DRAM speeds and non-volatility for enterprise storage. These innovations addressed challenges like endurance and power, with the embedded NVM market projected to reach $3.3 billion by 2030, fueled by eMRAM, ePCM, and eReRAM in MCUs and SoCs.

Electrically addressed non-volatile memories

Read-only and write-once devices

Read-only and write-once devices represent foundational forms of designed for permanent or single-use data storage, where the content is fixed during or immediately after manufacturing and cannot be altered thereafter. These technologies prioritize reliability for unchanging data such as firmware, boot code, and configuration parameters in , offering advantages in cost and density for high-volume production but sacrificing flexibility compared to erasable alternatives. Mask read-only memory (MROM), also known as masked ROM, is programmed with fixed data directly during the semiconductor fabrication process through custom masking and metallization layers that encode a permanent bit pattern into the chip's circuitry. This approach enables high-density storage at low per-unit cost, making MROM ideal for mass-produced applications where the data content is known in advance and does not require changes. For instance, MROM achieves densities in the range of megabits to gigabits per chip in advanced nodes, such as a 16-Mb implementation at 40 nm using dual trench isolation diode bitcells for efficient predefined data storage. Programmable read-only memory (PROM) extends the concept by allowing users to program the device once after manufacturing, typically by applying high-voltage pulses to blow fusible links or metal fuses that permanently alter the circuit paths. Invented in 1956 by Wen Tsing Chow at the for military applications like ICBM guidance, PROM became commercially available in the late 1960s and early 1970s, with early devices such as a 512-bit bipolar TTL version introduced by in 1970. This one-time programming capability provided flexibility for custom firmware without the need for full mask redesigns, though it still results in non-reversible content. One-time programmable (OTP) ROM serves as a modern variant of PROM, often integrated into standard CMOS processes using antifuse bit cells that are programmed via dielectric breakdown to create conductive paths, ensuring irreversible data fixation. OTP devices are particularly suited for storing secure keys, device identifiers, and calibration data in applications requiring tamper-resistant, non-alterable information, such as in microcontrollers and sensors. Examples include 1T or 2T antifuse cells developed since the early 2000s for embedded non-volatile storage without additional masks. These devices share key characteristics, including infinite read endurance due to their passive storage mechanisms that do not degrade with repeated access, the absence of any erase functionality, and non-volatility that retains data indefinitely without power. Densities can reach gigabit scales in contemporary implementations, supporting large firmware images, though their primary drawback is post-programming inflexibility, which limits use to scenarios with stable requirements. In modern contexts as of 2025, MROM, PROM, and OTP remain prevalent in cost-sensitive embedded systems, including microcontrollers for IoT devices and legacy game cartridges, where their simplicity and economy outweigh the need for rewritability.

Flash memory

Flash memory is a type of electrically addressed non-volatile memory that uses floating-gate transistors to store data as trapped charges, enabling repeated reprogramming unlike read-only devices. It dominates the non-volatile memory market, accounting for over 90% of shipments in 2025 due to its scalability and cost-effectiveness for mass storage. The technology was pioneered by at , who first described in a 1984 IEDM paper and in 1987, introducing the concept of block-erasable using . The core structure of flash memory relies on floating-gate metal-oxide-semiconductor field-effect transistors (FGMOSFETs), where a floating gate is isolated between the control gate and channel by insulating oxide layers, allowing charge storage to modulate the threshold voltage. NOR flash connects cells in parallel for random byte-level access, suitable for code execution, while NAND flash arranges cells in series for block-based operations, achieving higher density for data storage. This architectural difference makes NAND ideal for high-capacity applications, with densities scaling through multi-level cells. In 2025, advancements include SK hynix's mass production of 321-layer NAND flash. Programming and erasing in flash memory occur via Fowler-Nordheim tunneling, where high voltages (typically 15-20 V) enable quantum tunneling of electrons through the thin tunnel oxide into or out of the floating gate. The threshold voltage shift from injected charge is given by \Delta V_{th} = \frac{Q_{injected}}{C_{ox}} where Q_{injected} is the injected charge and C_{ox} is the oxide capacitance per unit area, determining the cell's logic state. Cells are categorized by bits per cell: single-level cells (SLC) store 1 bit for high endurance, multi-level cells (MLC) store 2 bits, triple-level cells (TLC) 3 bits, and quad-level cells (QLC) 4 bits, trading reliability for density in modern devices. To overcome planar scaling limits, 3D stacking emerged in the 2010s, vertically layering word lines and channels, with commercial products reaching 200+ layers by 2025, such as SK hynix's 321-layer NAND in mass production. This architecture boosts capacity while maintaining cell size, using techniques like charge-trap flash for better uniformity over floating gates. Performance includes read times of 10-100 μs, write/erase times around 1 ms, endurance of 10³-10⁵ program/erase cycles depending on cell type (e.g., 10⁵ for SLC, 10³ for QLC), and of at least 10 years at 55°C. Flash memory powers solid-state drives (SSDs) for computing storage and USB flash drives for portable data transfer, with NAND comprising the bulk due to its density. Challenges include limited endurance, addressed by wear-leveling algorithms that distribute writes evenly across blocks, and increasing error rates mitigated by low-density parity-check (LDPC) codes for error correction.

Ferroelectric RAM (FRAM)

Ferroelectric RAM (FRAM), also known as FeRAM, is a type of non-volatile random-access memory that leverages the ferroelectric properties of certain materials to store data. In ferroelectric materials, such as or , the application of an electric field induces a reversible polarization of the crystal lattice, creating two stable states that represent binary data: one for logic 0 and one for logic 1. This hysteresis effect allows the polarization to remain even after the field is removed, providing non-volatility without the need for constant power. The fundamental physics behind FRAM's data retention is described by the relationship between polarization P, electric field E, and material susceptibility \chi, given by the equation: P = \epsilon_0 \chi E where \epsilon_0 is the permittivity of free space. In ferroelectric materials, this relationship is nonlinear, exhibiting a hysteresis loop with a remnant polarization P_r that persists without an applied field, enabling stable data storage. FRAM employs a 1T-1C cell structure, consisting of one transistor and one ferroelectric capacitor per bit, similar to but with ferroelectric dielectrics replacing conventional oxides. During a write operation, an electric field is applied across the capacitor to align the polarization in the desired direction. Reads are destructive: sensing the charge on the capacitor disturbs the polarization state, necessitating an immediate write-back to restore the original data. Access times are typically around 50 ns, approaching speeds, while write cycles are fast and low-energy due to the minimal charge required to switch polarization. Endurance exceeds 10¹² cycles, far surpassing many other non-volatile memories. The concept of ferroelectric memory emerged in the 1950s with early research by organizations like Bell Labs and IBM exploring polarization effects in ceramics. Commercial development accelerated in the 1980s, culminating in the first FRAM chip from Ramtron International in 1993. By the 2000s, companies like Fujitsu integrated FRAM into microcontrollers for automotive and industrial uses. As of 2025, advancements include NXP's radiation-tolerant FRAM for aerospace applications and continued embedding in MCUs by RAMXEED (formerly Fujitsu Semiconductor Memory Solution), supporting densities up to 1 Mbit with enhanced reliability. FRAM's key advantages include operation at low voltages (1-3 V), ultra-low power consumption during writes, and inherent radiation hardness, making it suitable for embedded systems, smart meters, and space applications. However, challenges persist in scaling below 10 nm due to ferroelectric material fatigue and integration complexities, alongside higher fabrication costs compared to or .

Magnetoresistive RAM (MRAM)

Magnetoresistive random-access memory (MRAM) stores data using the magnetic states of ferromagnetic layers within a magnetic tunnel junction (MTJ), leveraging the tunneling magnetoresistance (TMR) effect for non-volatile retention. The MTJ consists of two ferromagnetic layers separated by a thin insulating barrier, typically magnesium oxide (MgO); one layer is fixed (pinned) while the other is free, allowing its magnetization direction to switch between parallel and antiparallel alignments relative to the fixed layer. In spin-transfer torque (STT) MRAM, the dominant modern variant, data writing occurs by passing a spin-polarized current through the MTJ, which exerts torque on the free layer's magnetization to flip its state without requiring an external magnetic field. The TMR effect quantifies the resistance difference between these states, defined as TMR = \frac{R_{AP} - R_P}{R_P} \times 100\%, where R_{AP} and R_P are the resistances in the antiparallel (high-resistance, logic '1') and parallel (low-resistance, logic '0') configurations, respectively, enabling reliable state detection. Reading in MRAM is non-destructive, achieved by applying a small sensing current to measure the MTJ's resistance change, distinguishing high and low states with high fidelity due to TMR ratios often exceeding 150%. Write operations in STT-MRAM typically require currents around 100 μA for sub-20 nm junctions, with access times of 10-35 ns, supporting high-speed embedded applications. Unlike charge-based memories, MRAM's magnetic storage avoids leakage currents, contributing to its low standby power. The development of MRAM began with a 1996 IBM prototype demonstrating field-induced switching in a 1 Mb array using early giant magnetoresistance principles, marking the first functional MRAM device. Commercialization advanced with Everspin Technologies' 2011 release of a 4 Mb standalone STT-MRAM product, transitioning from older toggle-mode designs to more scalable STT mechanisms. By 2025, TSMC has integrated 22 nm STT-MRAM into production for embedded applications, achieving densities up to 64 Mb with improved yield and reliability suitable for automotive and edge AI uses. STT-MRAM offers unlimited write endurance (>10^{15} cycles), data retention exceeding 10 years at elevated temperatures, and lower overall consumption compared to , making it ideal for last-level caches and persistent storage. However, challenges persist in scaling write currents as junction sizes shrink below 20 nm, potentially increasing error rates and energy demands without . Variants include the older toggle MRAM, which used magnetic fields for switching but suffered from higher and complexity, and emerging spin-orbit (SOT)-MRAM, which routes write currents parallel to the MTJ for reduced and faster operation in cache hierarchies.

Phase-change memory (PCM)

Phase-change memory (PCM) relies on the reversible phase transitions of chalcogenide materials, such as alloys in the Ge-Sb-Te () family, to store data non-volatily. These materials, like Ge₂Sb₂Te₅, exhibit two distinct states: an amorphous phase with high electrical resistance, typically representing the "0" state, and a crystalline phase with low resistance, representing the "1" state. The transition between these states is induced by controlled thermal processes, leveraging the material's ability to switch rapidly without mechanical movement. In operation, data writing involves via electrical current pulses applied through the memory cell. The "set" operation crystallizes the material by heating it to around 150–300°C for a sustained period (typically 10–100 ns), allowing atomic rearrangement into an ordered that reduces . The "reset" operation amorphizes the material by rapidly heating it above its (followed by quick to prevent recrystallization), increasing ; this also occurs on a nanosecond timescale. Reading is performed non-destructively by applying a low-voltage to measure the cell's , which differentiates the states by orders of magnitude. Access times for read/write operations range from 10–100 ns, with demonstrated endurance exceeding 10⁹ cycles per cell. As of 2025, has announced production-ready embedded PCM solutions for automotive and industrial microcontrollers. Key thermal parameters govern these transitions, including the melting temperature T_{melt} \approx 600^\circ \text{C} for Ge₂Sb₂Te₅, which sets the for amorphization. The current I_{reset} required for scales approximately as I_{reset} \propto \sqrt{\Delta T / \rho}, where \Delta T is the rise and \rho is the material's resistivity, highlighting the dependence on efficient localization to minimize power. The concept of PCM traces back to 1968, when Stanford Ovshinsky at Energy Conversion Devices demonstrated threshold switching in chalcogenide glasses, laying the foundation for phase-based memory. Commercial development accelerated in the 2000s with prototypes from and , including a 128 Mb embedded PCM chip in 2007 and scalable arrays demonstrating multi-level storage. and Micron's , introduced in 2015 as a hybrid using PCM material with an ovonic threshold selector for high-density stacking, represented a milestone, though the Optane product line was discontinued in 2022 due to market challenges; the underlying PCM technology persists in research and niche applications. PCM offers advantages in down to 10 nodes, enabling higher densities than some competing non-volatile technologies, and supports multi-bit cells by exploiting intermediate levels from partial . However, challenges include high write power demands due to requirements (often 10–100 µJ per bit) and drift in the amorphous state over time, which can degrade read margins without error correction.

Resistive RAM (ReRAM)

Resistive RAM (ReRAM), also known as , functions through resistive switching in thin metal oxide layers, where the device's changes between high and low s due to the formation and dissolution of conductive s primarily composed of oxygen vacancies. In change mechanism-based ReRAM, common materials include oxides such as oxide (TaOx) and oxide (HfOx), where the insulating oxide matrix transitions to a conductive path via localized oxygen vacancy aggregation. The process begins with a forming step, applying a sufficiently to initiate oxygen and create the initial , establishing the low (LRS). Subsequent operations involve a set process to reinforce the for LRS and a reset process to rupture it, restoring the high (HRS), all enabled by polarity-dependent voltage biases that drive electrochemical reactions at the interfaces. ReRAM switching is inherently voltage-driven, with applied modulating to control integrity, allowing for rapid state transitions. Access times can achieve sub-10 ns latencies, supporting high-speed read/write operations comparable to dynamic . Endurance typically spans 106 to 1012 cycles, varying by composition and electrode design, while energy consumption remains low at around 1–10 pJ per bit due to the localized nature of filamentary conduction. These operational traits position ReRAM as a candidate for embedded memory in energy-constrained systems. Early conceptualization of ReRAM traces to the 2000s, with Laboratories and demonstrating prototype memristive switches in titanium dioxide thin films, laying groundwork for filament-based resistive effects. By 2012, commercialized initial conductive bridging variants, focusing on scalable integration. achieved the first mass production of ReRAM-embedded microcontrollers in 2013, targeting low-power portable devices. Entering 2025, continues production scaling, while Weebit Nano has advanced to full manufacturing qualification, emphasizing crossbar array architectures for neuromorphic hardware that mimic through analog resistance tuning. Key advantages of ReRAM include seamless compatibility with standard fabrication processes, enabling backend-of-line integration without altering front-end logic flows, and inherent scalability to stacking, which supports vertical crossbar arrays for densities exceeding 1 Tb/cm². However, drawbacks persist, such as cycle-to-cycle variability in switching voltages and resistance levels arising from stochastic filament formation, alongside the need for an initial forming voltage—often 3–5 V—to nucleate the first , which can stress device reliability and increase power overhead during initialization. In the filamentary conduction model, the R of the low- state is described by R = \rho \frac{L}{A} where \rho is the material resistivity, L the length (typically spanning the oxide thickness), and A the effective cross-sectional area of the vacancy cluster; switching primarily modulates A through vacancy changes. During forming and set, a current limit—enforced via external circuitry—prevents excessive current flow, controlling radius and averting breakdown. A prominent variant is conductive bridging RAM (CBRAM), which employs electrochemical mechanisms where metal ions from an active electrode, such as copper (Cu), dissolve and electrodeposit to form metallic rather than vacancy-based filaments, often in solid electrolytes like chalcogenides.

Field-effect transistor-based memories (FeFET)

Field-effect transistor-based memories, known as FeFETs, integrate a ferroelectric material directly into the gate stack of a metal-oxide-semiconductor field-effect transistor (MOSFET). The core principle relies on the ferroelectric layer's ability to generate a non-volatile remnant polarization that modulates the transistor's threshold voltage (V_th). When polarized in one direction, the bound charges at the ferroelectric interfaces induce a shift in the channel potential, altering V_th and thereby controlling the drain current without power. This enables binary or multi-level data storage, where the two polarization states correspond to distinct V_th levels, providing non-volatility through the hysteresis in the ferroelectric material. Operation of FeFETs involves writing data by applying a poling across the to switch the , which induces a reversible V_th shift (ΔV_th). The can be expressed as V_{th} = V_{th0} \pm \Delta V_p, where V_{th0} is the initial without , and \Delta V_p = \frac{3}{2} \cdot \frac{P_r \cdot t_{fe}}{\varepsilon_{fe}}, with P_r as the remnant , t_{fe} as the ferroelectric layer thickness, and \varepsilon_{fe} as the . Reading occurs non-destructively by applying a sense voltage and measuring the drain current, which reflects the polarized state without disturbing it. Access times are typically around 10-25 ns for read operations, with write times under 1 μs, and endurance can exceed 10^7 cycles in oxide-based devices, though optimized structures achieve up to 10^10 cycles. Research on FeFETs originated in the 1990s, focusing on perovskite materials like (PZT) for gate integration, though challenges with scalability and compatibility persisted. A major advancement came in 2017 when demonstrated hafnium oxide (HfO2)-based FeFETs, leveraging doped HfO2's ferroelectric properties discovered through high-temperature annealing, enabling seamless integration with existing processes. By 2025, FeFETs have been demonstrated in 28 nm high-k (HKMG) processes for potential embedded applications, particularly for accelerators and in-memory , where multi-bit cells support efficient synaptic weights in neuromorphic systems. FeFETs offer key advantages including fast switching speeds comparable to , low operating voltages (under 3-5 V), and high compatibility with logic circuits, facilitating dense non-volatile memory with cell areas as small as 10 F². Their low power consumption (10-15 fJ/bit) and scalability to finFET or gate-all-around architectures make them suitable for edge AI and . However, drawbacks include limited of 1-10 years due to depolarization fields and charge at interfaces, as well as ferroelectric , which degrades endurance over repeated cycles and introduces variability in V_th shifts. In comparison to ferroelectric RAM (FRAM), which relies on a 1T-1C structure with a separate ferroelectric , FeFETs use a single-transistor cell for higher density and simpler integration but face trade-offs in endurance and retention.

Mechanically addressed systems

Magnetic storage systems

systems store data by magnetizing microscopic domains on a , such as rotating in hard disk drives (HDDs) or linear tape media, where binary states (0 or 1) are represented by the orientation of . Read and write operations are performed using specialized heads: inductive heads generate to align domains during writing, while magnetoresistive (MR) heads detect changes in for reading by exploiting the effect, where electrical resistance varies with the applied . This enables non-volatile retention of data without power, as the magnetic orientations persist indefinitely. Hard disk drives operate on spinning aluminum or platters coated with ferromagnetic , typically rotating at 5,400 to 7,200 RPM, with read/write heads mounted on arms that position over concentric tracks. The first commercial HDD, IBM's RAMAC 305 system introduced in 1956, featured 50 24-inch platters storing 5 MB total, marking the advent of random-access . By 2025, advancements like (HAMR), commercially available since mid-2025, have enabled areal densities exceeding 1 Tb/in², allowing single-drive capacities of 36 TB or more, as seen in Seagate's Exos M series. Typical seek times for these drives range from 5 to 10 ms, balancing mechanical positioning with high-throughput . Magnetic tape systems employ linear serpentine recording, where data is written in parallel tracks across the tape width as it moves past stationary heads, reversing direction for multiple passes to maximize capacity. The LTO-10 format, the current specification as of November 2025, provides 40 TB native capacity per cartridge (100 TB compressed at 2.5:1 ratio), with drives available since mid-2025 and higher-capacity media shipping in early 2026; it is optimized for long-term archival storage with shelf lives exceeding 30 years under controlled conditions. These tapes are housed in robotic libraries for bulk data management, offering cost-effective scalability for petabyte-scale repositories. Magnetic storage excels in providing high capacities at low cost per gigabyte—often under $0.02/GB for HDDs—making it ideal for secondary storage in data centers and consumer devices, though it suffers from mechanical vulnerabilities like head crashes and slower random access compared to semiconductor non-volatile memory.

Optical storage systems

Optical storage systems store data mechanically using laser light to read variations in reflectivity on a rotating disc, typically made of a polycarbonate substrate. Information is encoded as microscopic pits and lands etched into a spiral track on the substrate's surface, where pits are depressions that scatter laser light, reducing reflection, while lands are flat areas that reflect it strongly. A thin reflective metal layer, often aluminum, covers the pits and lands, and a protective lacquer seals the structure. This read-only principle ensures non-volatility, as data persists without power, relying on the physical permanence of the etched features. Key types include read-only formats like the , introduced in 1982 with a capacity of 650 MB, suitable for audio and ; the DVD-ROM, standardized in 1995 offering 4.7 GB per single-layer side for video and data; and Blu-ray Disc (BD), released in 2006 with 25 GB single-layer capacity, scaling to 100 GB in multi-layer variants like BD-XL triple-layer discs. Write-once variants, such as and DVD-R, allow permanent recording by burning a layer to mimic pits, while rewritable types like and DVD-RW employ phase-change materials that switch between crystalline (reflective) and amorphous (absorptive) states via laser heating, enabling data overwriting. This phase-change mechanism shares conceptual similarity with phase-change memory but operates in a mechanical disc format. Operation involves a low-power directing a through an objective lens onto the spinning , with a detecting reflected light to distinguish from lands as and 1s. Wavelengths progress from 780 nm for , enabling larger around 0.83 μm long, to 650 nm for DVDs (pit length ~0.4 μm), and 405 nm for Blu-ray (pit length ~0.16 μm), allowing denser packing. Access times average 100-200 due to seeking of the laser head, far slower than solid-state alternatives. Rewritable discs endure 10^3 to 10^5 cycles before material degradation reduces reliability. The minimum resolvable feature size, or pit length, is governed by the diffraction limit: d = \frac{\lambda}{2 \cdot \mathrm{NA}} where d is the resolution, \lambda is the laser wavelength, and NA is the numerical aperture of the objective lens (typically 0.45-0.85). As of , optical storage has declined in consumer applications, supplanted by solid-state drives and cloud services, but persists in archival roles for its longevity exceeding decades under proper storage. Multi-layer Blu-ray variants reach 100 GB, supporting cold in enterprises. Advantages include easy removability for transport and offline , plus inherent tamper resistance in read-only or write-once formats, where altering data requires physical replacement. Drawbacks encompass slower access and transfer rates (up to 16x for Blu-ray, or ~72 MB/s) compared to hard disk drives, alongside ceilings limiting beyond archival niches.

Organic and emerging non-volatile memories

Organic memory devices

Organic memory devices utilize , such as pentacene, in resistive or ferroelectric configurations to achieve non-volatile through mechanisms like charge trapping or ferroelectric . These devices operate by exploiting the ability of organic materials to maintain distinct high- and low-conductance states under applied , enabling bistable or multilevel switching without power consumption for retention. Key types include organic resistive random access memory (RRAM) employing dielectrics for filamentary or interface-based switching, and organic ferroelectric RAM () based on (PVDF) polymers, where switching via ion migration or alignment stores information. In organic RRAM, charge carriers form conductive paths in matrices under bias, while organic leverages the remnant in PVDF copolymers like P(VDF-TrFE) for hysteresis-based effects. Performance characteristics typically include access times in the range (1-100 ms), endurance of 10^3 to 10^5 cycles, and exceeding 10^5 seconds (with some extrapolated to over 10 years), facilitated by low-temperature below 200°C compatible with flexible substrates. These metrics support applications in , though variability arises from material morphology and environmental factors. As of October 2025, organic memristors are advancing toward brain-inspired applications in flexible wearables for . Research on organic memory devices emerged in the , with seminal reviews highlighting early switching behaviors in organic thin films. By 2025, prototypes integrated into wearables demonstrate feasibility for flexible, on-body computing, but commercial remains limited due to challenges. Advantages encompass mechanical flexibility for bendable , biocompatibility for biomedical uses, and low-cost solution processing, contrasting with rigid inorganic alternatives. Drawbacks include environmental issues like from or oxygen, and low integration below 1 Mb/cm² owing to larger feature sizes in fabrication. Charge transport in these devices is governed by carrier mobility, expressed as \mu = \frac{q \tau}{m^*} where q is the , \tau the mean time, and m^* the effective mass; in , \mu is typically low at approximately $10^{-3} cm²/Vs due to disorder-induced .

Advanced persistent memory technologies

Advanced technologies represent cutting-edge developments in non-volatile memory (NVM) that extend beyond conventional types like phase-change memory (PCM) and resistive RAM (ReRAM) to enable byte-addressable persistence and hybrid architectures for data centers and workloads. These innovations aim to provide DRAM-like speed with non-volatility, addressing the performance gap between volatile main memory and slower secondary storage. Key examples include , which combined PCM elements with a chalcogenide selector in a cross-point to achieve storage class memory (SCM) capabilities, offering latencies closer to DRAM while maintaining persistence. Although discontinued by Micron in 2021 and in 2022 due to economic challenges, 3D XPoint proved influential in demonstrating scalable, high-density NVM for enterprise applications, paving the way for subsequent SCM solutions. Emerging research in the has explored single-molecule memories as a path to ultimate scaling, leveraging -active molecules to switch between bistable or multistable states for at the level. These devices utilize mechanisms such as charge and conformational changes to achieve non-volatile retention, with potential densities far exceeding traditional silicon-based NVM due to their molecular-scale . A 2024 review highlights progress in integrating these into functional prototypes, emphasizing states for reliable switching and addressing challenges like with existing circuitry for practical deployment. Compute-in-memory architectures further advance NVM by embedding computation directly within memory arrays, reducing data movement bottlenecks in inference and . ReRAM-based arrays, for instance, enable analog matrix-vector multiplications for neural networks, with recent 2025 studies demonstrating energy-efficient designs using 1T2R configurations that achieve high accuracy in tasks like element-wise operations. These systems, often scaled below 10 nm, support edge by combining persistence with in-situ processing, though they face hurdles in precision and variability. In 2025 trends, (CXL) is enabling pooled across and platforms, allowing disaggregated NVM modules to act as scalable extensions for data centers. Demonstrations at events like FMS 2025 showcase CXL Type-3 devices with processors providing high-performance checkpointing, outperforming traditional SSDs in recovery times for large-scale simulations. Samsung's advancements in spin-transfer torque MRAM (STT-MRAM) for applications, including 14 nm eMRAM entering mass production by late 2024 for and chips, further support low-power persistence in LPDDR-integrated SoCs. The market is projected to grow at a 10.1% CAGR from 2025 to 2035, driven by demand for hybrid DRAM-NVM systems. These technologies bridge the DRAM-SSD performance gap by offering byte-addressable access and crash consistency, but challenges persist in high costs and ecosystem maturity for widespread adoption.

Applications

Secondary storage and consumer devices

Solid-state drives (SSDs) have become the dominant form of secondary storage in consumer devices, primarily utilizing NAND flash memory for its high density and non-volatility. The NVMe interface, optimized for flash-based storage, enables significantly faster data transfer rates compared to traditional connections, making SSDs ideal for operating system boot times and application loading in personal computers. By 2025, consumer SSDs routinely offer capacities up to 8 TB, with PCIe 5.0 models achieving sequential read and write speeds exceeding 10 GB/s, such as the 9100 PRO series reaching 14,800 MB/s reads and 13,400 MB/s writes. Portable storage solutions like USB flash drives and SD cards also rely heavily on NAND flash, often incorporating hybrid architectures that combine NAND for bulk data storage with NOR flash for faster random access in boot or code execution scenarios. These devices provide convenient, removable non-volatile storage for consumers, with USB drives supporting capacities up to 2 TB in compact form factors suitable for backups and file transfers. Similarly, SD cards, widely used in cameras and portable media players, achieve up to 2 TB capacities, enabling high-resolution video recording and large photo libraries without frequent data offloading. In , non-volatile memory powers essential storage in smartphones, digital cameras, and tablets, where embedded (eMMC) and (UFS) standards based on deliver reliable performance for apps, , and . Smartphones increasingly adopt UFS for its superior speed over eMMC, supporting multitasking and 4K video storage in devices with 256 GB to 1 TB capacities. The global market, driven largely by demand from these applications, exceeded $70 billion in 2025, reflecting its critical role in the portable electronics ecosystem. While solid-state non-volatile memory dominates modern secondary storage, mechanical systems like hard disk drives (HDDs) continue to serve complementary roles in desktops and laptops for cost-effective, high-capacity bulk storage, often reaching 20 TB or more per drive. remains vital for long-term backups and archival in consumer and setups, offering terabyte-scale capacities at low cost per with offline security against . To enhance data protection in these non-volatile systems, features like wear-leveling distribute write operations evenly across cells to prevent premature failure, while TCG encryption standards enable hardware-based AES-256 self-encrypting drives that secure without performance overhead.

Non-volatile main memory

Non-volatile main memory refers to technologies that provide persistent, byte-addressable storage functioning as system RAM, retaining data without power while offering DRAM-like access speeds. Key technologies include magnetoresistive random-access memory (MRAM) and phase-change memory (PCM), which leverage magnetic tunneling junctions and chalcogenide phase transitions, respectively, to enable non-volatility in DIMM form factors. By 2025, advancements such as Samsung's development of high-density MRAM chips targeting 16 Gb densities have progressed toward integration in server DIMMs, while concepts for successors to Intel's Optane PCM-based persistent memory, including Samsung's revived Z-NAND and Solidigm's storage-class memory solutions, aim to bridge DRAM and storage hierarchies with enhanced endurance and speed. These technologies enable significant benefits in systems, particularly instant-on capabilities that eliminate traditional times by preserving application states across power cycles, and rapid recovery through persistence without filesystem flushes. Interfaces such as NVDIMM-P, which pairs with non-volatile backing via the DDR bus, and the (CXL) 2.0 standard—ratified and widely adopted by 2025—facilitate disaggregated, poolable in centers, allowing non-volatile modules to extend capacity beyond limits while maintaining cache-coherency. In server environments, this supports high-performance workloads like in- databases, where non-volatile main reduces for frequent reads and ensures durability during failures. Performance characteristics of non-volatile main memory typically include read latencies of 50-200 ns and write latencies approaching levels in optimized designs, with exceeding 100 GB/s in multi-channel configurations, making it suitable for latency-sensitive applications despite modest penalties compared to pure volatile . systems often employ caching strategies to mitigate these differences, where effective can be modeled as
L_{\text{eff}} = L_{\text{NVM}} \times (1 - \text{hit\_rate}) + L_{\text{DRAM}} \times \text{hit\_rate}
balancing non-volatile with volatile speed for overall system efficiency.
Despite these advantages, challenges persist, including costs approximately 10 times higher than per bit due to complex fabrication processes, limiting widespread adoption to high-value enterprise use cases. Volatility trade-offs, such as higher write latencies and limits in PCM (around 10^8-10^9 cycles), require architectural mitigations like wear-leveling. The non-volatile memory market, encompassing main memory applications, is projected to exceed $1 billion in 2025, driven by server and deployments, though scaling production remains key to cost reduction.

Embedded and specialized systems

In embedded systems, non-volatile memory (NVM) plays a critical role in resource-constrained environments such as microcontrollers (MCUs), where it enables persistent without the power overhead of traditional volatile alternatives. Ferroelectric (FRAM) and (ReRAM) are prominent embedded NVM (eNVM) technologies integrated into MCUs for applications requiring low-power operation and high reliability, such as intermittently powered devices that harvest energy from the environment. For instance, serves as main in certain MCUs due to its compatibility with non-volatile needs in sensors, while ReRAM offers scalability and fast switching for on-chip in compact systems. These technologies provide advantages over conventional , with and MRAM supporting up to 10^12 write cycles in some configurations, far exceeding flash's typical 10^5 limits. The market for embedded emerging NVM is experiencing rapid growth, driven by demand in MCUs and system-on-chips (SoCs) for . Projections indicate that the embedded emerging NVM segment, including eMRAM, ePCM, and eReRAM, will reach $3.3 billion by 2030, reflecting adoption in next-generation devices. A key advancement is TSMC's 16nm FinFET embedded MRAM (eMRAM), which has achieved production readiness and supports high-density integration for automotive and consumer MCUs, offering 1 million write cycles and 20-year at elevated temperatures. In automotive applications, NVM must withstand harsh conditions, leading to widespread use of AEC-Q100 qualified MRAM and in electronic control units (ECUs). These devices operate reliably across a temperature range of -40°C to 125°C, ensuring in engine management and advanced driver-assistance systems (ADAS). For example, Everspin's MRAM solutions meet Grade 1 AEC-Q100 standards, providing non-volatile storage with unlimited endurance for real-time ECU updates, while variants from Winbond support over-the-air reprogramming in vehicle infotainment. Aerospace and systems demand radiation-hardened NVM to maintain functionality in high-radiation environments, where and MRAM excel due to their inherent to single-event upsets and total ionizing dose effects. Radiation-hard from Infineon is qualified for under QML-V standards, offering non-volatility and retention exceeding 10 years in without power. Similarly, MRAM devices from and Everspin provide rad-hard solutions for satellites and , with intrinsic immunity to cosmic rays ensuring data retention during prolonged missions. In medical devices, biocompatible NVM prototypes are emerging to enable implantable that safely with biological tissues. These devices, often based on organic memristors, demonstrate non-volatile resistive switching using natural materials like extracted from peels, achieving eco-friendly, flexible memory suitable for bio. Bioresorbable organic-inorganic hybrids further support long-term neural implants, with prototypes showing multilevel storage and for applications without eliciting immune responses. Current trends in embedded NVM emphasize integration with edge through in-memory computing architectures, which minimize data movement to achieve ultra-low power consumption. These systems leverage eNVM like ReRAM for on-chip , enabling below 1 μW while supporting in battery-limited devices such as wearables and sensors.

References

  1. [1]
    What are non-volatile memories and solid-state drives?
    Non-volatile memory (NVM) or non-volatile storage is a type of computer memory that can retain stored information even after power is removed.
  2. [2]
    [PDF] Memory
    All memory types can be categorized as ROM or RAM, and as volatile or non-volatile: • Read-Only Memory (ROM) cannot be modified (written), as the name implies.
  3. [3]
    Emerging Nonvolatile Memory Technologies in the Future of ...
    Jun 29, 2025 · Research on nonvolatile memories began with the development of charge memory devices in the 1960s.<|control11|><|separator|>
  4. [4]
  5. [5]
    Pioneers of Semiconductor Non-Volatile Memory (NVM): The First ...
    May 18, 2020 · George Perlegos contributed to several important NVM developments at Intel, including the company's first EEPROM, the 2816, 16K-bit device in ...
  6. [6]
    Overview of emerging nonvolatile memory technologies - PMC - NIH
    A variant of charge storage memory referred to as Flash memory is widely used in consumer electronic products such as cell phones and music players while NAND ...
  7. [7]
    [PDF] Novel Non-Volatile Memory Devices and Applications
    May 11, 2023 · Today, solid-state memory devices are ubiquitous and used in a wide range of applications, including personal computers, smartphones, wearable ...
  8. [8]
  9. [9]
    2.3.2. Non-Volatile Memory - Intel
    Non-volatile memory commonly stores processor boot-code, persistent application settings, and Intel FPGA configuration data. Although non-volatile memory has ...
  10. [10]
    Non-Volatile Memory - an overview | ScienceDirect Topics
    Example applications of both types of flash memory include personal computers and all kinds of embedded systems such as digital audio players, digital ...Types and Technologies of... · Integration of Non-Volatile...
  11. [11]
    Review on Non-Volatile Memory with High-k Dielectrics - NIH
    The ability to retain charge without supplied power is the definition of non-volatile memory. Retention is a metric used to quantify the extent of time expected ...
  12. [12]
    Attack of the Killer Microseconds - Communications of the ACM
    Apr 1, 2017 · Likewise, raw flash device latency is on the order of tens of microseconds. The latencies of emerging new non-volatile memory technologies ...
  13. [13]
    Millipede memory - Wikipedia
    Millipede memory is a form of non-volatile computer memory. It promised a data density of more than 1 terabit per square inch (1 gigabit per square millimeter)
  14. [14]
    Ferroelectric RAM - Wikipedia
    ... endurance (about 1010 to 1015 cycles). FeRAMs have data retention times of more than 10 years at +85 °C (up to many decades at lower temperatures).
  15. [15]
    batteries - Flash memory power consumption
    May 1, 2014 · A flash chip requires power for either read or write access. The power advantage of flash comes from it being non-volatile. That is, it needs no power.
  16. [16]
    [PDF] Chapter 5 Amorphous silicon TFT based non-volatile memory
    Retention: the amount of time the element can hold its current state without significant changes. Endurance: the number of cycles of write/erase the element can ...
  17. [17]
    Phase change materials in non-volatile storage - ScienceDirect.com
    Phase change materials are often semiconducting or semi-metallic alloys containing the elements of group VI of the periodic table, excluding oxygen.
  18. [18]
    Bi-exponential decay model of electron in Al2O3/Au ... - AIP Publishing
    Sep 21, 2022 · A common non-volatile memory device has a multi-stacking structure ... retention behavior was confirmed by fitting a mono-exponential function.
  19. [19]
    Modeling charge variation during data retention of MLC Flash ...
    In this paper, we propose to model charge variation in Multi-Level Cells in NOR Flash memories. We first define a sensitivity-to-temperature factor to ...Missing: decay | Show results with:decay
  20. [20]
    1801: Punched cards control Jacquard loom | The Storage Engine
    In 1801, Joseph Jacquard used punched cards to control a loom, enabling complex patterns. Later, punched cards were used for data storage and input.
  21. [21]
    The Jacquard Loom: A Driver of the Industrial Revolution
    Jan 1, 2019 · The Jacquard loom was the first to automatically create complex textile patterns using punched cards, enabling mass production of intricate ...
  22. [22]
    Paper Tape - an overview | ScienceDirect Topics
    In 1857, Sir Charles Wheatstone introduced the first application of paper tapes as a medium for the preparation, storage, and transmission of data.Missing: retention | Show results with:retention
  23. [23]
    1898: Poulsen records voice on magnetic wire | The Storage Engine
    In 1898 Danish inventor Valdemar Poulsen (1869–1942) recorded his voice by feeding a telephone microphone signal to an electromagnet that he moved along a ...
  24. [24]
    1935: Audio recorder uses low-cost magnetic tape
    Austro-German engineer Fritz Pfleumer (1881 – 1945) coated 16 mm wide paper strips with fine granules of iron powder as a medium for magnetic recording. He ...
  25. [25]
    1932: Tauschek patents magnetic drum storage
    Nov 27, 2015 · Austrian engineer Gustav Tauschek (1899-1945) demonstrated and patented a prototype magnetic drum storage device.
  26. [26]
    Memory & Storage | Timeline of Computer History
    Magnetic tape allows for inexpensive mass storage of information and is a key part of the computer revolution. The IBM 726 was an early and important practical ...Missing: century | Show results with:century
  27. [27]
    NIHF Inductee An Wang Invented Magnetic-Core Memory
    He founded Wang Laboratories in 1951 to develop specialty electronic devices. In 1965 he introduced LOCI, a desk top computer. This forerunner of the Wang ...
  28. [28]
    1956: First commercial hard disk drive shipped | The Storage Engine
    IBM developed and shipped the first commercial Hard Disk Drive (HDD), the Model 350 disk storage unit, to Zellerbach Paper, San Francisco in June 1956.
  29. [29]
    Magnetic Tape - Engineering and Technology History Wiki
    Apr 1, 2019 · In 1928 Fritz Pfleumer developed, and in 1929 patented a magnetic recording tape using oxide bonded to a strip of paper or film. Building ...
  30. [30]
    [PDF] evolution of nonvolatile semiconductor memory - CERN Indico
    The Transistor (1947) was not just a replacement for vacuum tubes, it created the modern computer and the new industrial revolution.
  31. [31]
    A Brief History of the MOS transistor, Part 1: Early Visionaries
    Apr 3, 2023 · Along with colleague Simon Min Sze, Kahng developed the floating-gate MOSFET in 1967. This invention is the core storage element used in EPROMs ...
  32. [32]
    1960: Metal Oxide Semiconductor (MOS) Transistor Demonstrated
    John Atalla and Dawon Kahng fabricate working transistors and demonstrate the first successful MOS field-effect amplifier. Figure from Dawon Kahang's MOS patent.
  33. [33]
    1971: Reusable semiconductor ROM introduced | The Storage Engine
    Early integrated circuit ROMs offered a non-volatile form of semiconductor memory (NVM) but required a custom mask for each design. In 1970, Radiation Inc.
  34. [34]
    Evolution of Non-Volatile Memory: Eli Harari's system level floating ...
    Mar 18, 2017 · Harari's invention of EEPROM non-volatile memory kicked off a decade of major discoveries in solid-state memory for computing devices.Missing: history 1960s 1970s ROM<|separator|>
  35. [35]
    Chip Hall of Fame: Toshiba NAND Flash Memory - IEEE Spectrum
    Sep 28, 2025 · The saga that is the invention of flash memory began when a Toshiba factory manager named Fujio Masuoka decided he'd reinvent semiconductor memory.
  36. [36]
    Understand the Importance of NAND Flash as KIOXIA Celebrates 35 ...
    Feb 10, 2022 · It is NAND flash and NAND was invented by Toshiba, now KIOXIA, way back in 1987 after being first discovered by Dr. Fujio Masuoka.
  37. [37]
    Ferroelectric field effect transistors for electronics and optoelectronics
    Feb 23, 2023 · In 1993, the company RAMTRON successfully developed the first FRAM products.153 Thereafter, some major semiconductor companies, such as Fujitsu, ...
  38. [38]
    Advances in Spin-Transfer-Torque MRAM for MRS Fall Meeting 2022
    Nov 27, 2022 · IBM has had a long history of research and development of STT-MRAM. Spin-transfer-torque switching was invented at IBM in 1996 by John ...Missing: prototype | Show results with:prototype
  39. [39]
    Phase-Change Memory for In-Memory Computing - PMC
    In 2011, IBM demonstrated 512-Mbit 2bit/cell multilevel memory cells 90 nm, followed by 4bit/cell in 2013. A significant advancement in PCM technology occurred ...Missing: milestone | Show results with:milestone
  40. [40]
    What Happened To ReRAM? - Semiconductor Engineering
    Sep 21, 2017 · In the works for years, ReRAM gained notoriety in 2008, when Hewlett-Packard observed one type of ReRAM, dubbed the memristor. For years, HP ...
  41. [41]
    STT-MRAM: Introduction and market status
    Jul 3, 2024 · Volume production is expected "soon". In August 2016 Everspin started sampling pMTJ-based ST-MRAM chips.
  42. [42]
    Emerging Non-Volatile Memory 2025 - Yole Group
    Sep 11, 2025 · Embedded emerging NVMs are set to expand to $3.3B by 2030, driven by the rapid rise of eMRAM, ePCM, and eRRAM in next-generation MCUs and SoCs.
  43. [43]
    Read Only Memory - an overview | ScienceDirect Topics
    Masked ROM (MROM) is programmed during the manufacturing process using masking and metallization, resulting in a permanent bit pattern that cannot be changed.
  44. [44]
    A 40-nm 16-Mb Contact-Programming Mask ROM Using Dual ...
    Aug 7, 2025 · Read Only Memories (ROMs) serve as important non-volatile memory in various hardware systems to store predefined data and programs, which is ...
  45. [45]
    [PDF] 9 rom, eprom, and eeprom technology
    EPROMs were created in the 1970s and have long been the cornerstone of the non-volatile memory market. But the development of flash memory devices (see Section ...
  46. [46]
    One-Time-Programmable Memory (OTP) - Semiconductor Engineering
    OTP memory can be written to once, programmed after manufacturing, and its contents cannot be changed after being programmed.Missing: calibration | Show results with:calibration
  47. [47]
  48. [48]
    25 Microchips That Shook the World - IEEE Spectrum
    May 1, 2009 · The saga that is the invention of flash memory began when a Toshiba factory manager named Fujio Masuoka decided he'd reinvent semiconductor ...
  49. [49]
    Status of the Memory Industry 2025 - Yole Group
    Jun 5, 2025 · Memory industry reinvented by AI and HBM, overcoming cyclicality and poised for explosive growth, reaching $200B by 2025.
  50. [50]
    Moore's Law Milestones - IEEE Spectrum
    Apr 30, 2015 · At the meeting, Fujio Masuoka describes what is now called NOR flash. A few years later, at the 1987 IEDM, Masuoka describes the serial NAND ...
  51. [51]
    2.1.1 Flash Memory - IuE
    2.1.1.2 NOR. In NOR gate flash memory each cell consists of a standard MOSFET with two gates instead of one. The top gate is the so called Control Gate (CG) ...
  52. [52]
    What is NAND Flash memory/ NAND versus NOR
    A flash cell is effectively an N-channel MOSFET transistor with an extra gate – called the floating gate – which is sandwiched between two insulators. Probably ...
  53. [53]
    The Invention of NAND Flash Memory - 3D InCites
    Mar 21, 2018 · More than 20 years after D. Kahng and S.M. Sze's floating gate concept, the application of Fowler-Nordheim tunneling for both program and erase ...<|separator|>
  54. [54]
    [PDF] Analysis of Radiation Effect on the Threshold Voltage of Flash ...
    It should be noted that ∆VTH equation of FGT is completely different from charge trap memory, which explained in [1], because of the structural and charge ...Missing: Q_injected / C_ox
  55. [55]
    A Guide to NAND Flash Memory - SLC, MLC, TLC, and QLC - SSSTC
    Understanding NAND Flash Types: SLC, MLC, TLC, and QLC · What is NAND Flash Memory? · Single-Level Cell (SLC) NAND Flash · Multi-Level Cell (MLC) NAND Flash.
  56. [56]
    One-Team Spirit: SK hynix's 321-Layer NAND
    Jul 7, 2025 · For NAND flash products, the storage capacity depends on the number of stacked cell1 layers. Therefore, increasing the number of these layers is ...
  57. [57]
    [PDF] LDPC-in-SSD: Making Advanced Error Correction Codes ... - USENIX
    Abstract. Conventional error correction codes (ECCs), such as the commonly used BCH code, have become increasingly i- nadequate for solid state drives ...
  58. [58]
    [PDF] NAND Flash Memory Reliability in Embedded Computer Systems
    Data retention is a function of the rate of charge loss and the amount of margin between charge regions. The charge loss is strongly driven by the temperature ...Missing: decay | Show results with:decay
  59. [59]
    Difference between SLC, MLC, TLC and 3D NAND in USB flash ...
    Common types of NAND flash storage are SLC, MLC, TLC and 3D NAND. This article discusses the different characteristics of each type of NAND.
  60. [60]
    LDPC ECC in SSDs: How Error Correction Protects Data
    Jun 6, 2019 · Error correction codes (ECC) are techniques used to correct errors in the NAND flash, allowing recovery of data that may be corrupted due to bit errors.
  61. [61]
    F-RAM (Ferroelectric RAM) - Infineon Technologies
    F-RAM memories are built on ferroelectric technology. The F-RAM chip contains a thin ferroelectric film of lead zirconate titanate, commonly referred to as PZT.
  62. [62]
    FRAM Ferroelectric RAM Technology & Operation - Electronics Notes
    Ferroelectric RAM operation and technology is based upon the properties crystals of a dielectric that have a reversible electric polarisation. FRAM memory ...
  63. [63]
    Memory 101: What you need to know about FRAM, part 1 - EE Times
    Feb 12, 2013 · Embedding ferroelectric capacitors into an integrated circuit results in FRAM. FRAM operates internally like DRAM but has the high speed of SRAM ...
  64. [64]
    (PDF) General Introduction to Ferroelectrics - ResearchGate
    ... polarization (P)is. directly proportional to the macroscopic electric field EðÞat a given point of a. dielectric [5–15]. It is expressed as. P∝E)P¼ε0χeE(3).
  65. [65]
  66. [66]
    Inside a ferroelectric RAM chip - Ken Shirriff's blog
    Ramtron was acquired by Cypress Semiconductor in 2012 and then Cypress was acquired by Infineon in 2019. Infineon still sells FRAM, but it is a niche product, ...Missing: commercialization source
  67. [67]
    Ferroelectric RAM Market Size, Share and Growth Report 2032
    Aug 21, 2025 · Recent Developments. In May 2025, NXP launched a new FeRAM product with enhanced radiation tolerance for aerospace applications, achieving ...
  68. [68]
    FUJITSU Semiconductor Memory Becomes RAMXEED - GLYN
    RAMXEED Limited, formerly Fujitsu Semiconductor Memory Solution, is a leading developer of innovative memory technologies. The name change on 1 January 2025 ...
  69. [69]
    [PDF] A survey of circuit innovations in ferroelectric random-access ...
    This paper surveys circuit innovations in ferroelectric memo- ries at three circuit levels: memory cell, sensing, and architecture.
  70. [70]
    [PDF] Scaling Projections on Spin Transfer Torque Magnetic Tunnel ...
    Dec 12, 2017 · and is measured by tunnel magnetoresistance (TMR) defined as. TMR = (RAP − RP )/RP , where RAP and RP are the resis- tances of the device in ...
  71. [71]
    Recent progress in spin-orbit torque magnetic random-access memory
    Oct 1, 2024 · Shrinking the MTJ size to ~20 nm while maintaining retention requirements could reduce switching current to ~100 μA. This scaling approach ...
  72. [72]
    [PDF] Evaluation of STT-MRAM as a Scratchpad for Training in ML ... - arXiv
    Aug 3, 2023 · In contrast, STT-MRAM provides the highest endurance (> 1015 cycles) with 1 to 10 ns access time [14] while also providing 3-4X higher density ...
  73. [73]
    MRAM technologies: from space applications to unified cache ...
    Sep 29, 2021 · MRAM is a non-volatile memory technology that relies on the (relative) magnetization state of two ferromagnetic layers to store binary information.
  74. [74]
    Researchers celebrate 20th anniversary of IBM's invention of Spin ...
    Jul 6, 2016 · Researchers celebrate 20th anniversary of IBM's invention of Spin Torque MRAM by demonstrating scalability for the next decade.Missing: prototype | Show results with:prototype
  75. [75]
    MRAM history
    May 25, 2025 · The early years. 1989 - IBM scientists made a string of key discoveries about the "giant magnetoresistive" effect in thin-film structures.
  76. [76]
    TSMC - MRAM-Info
    The company announced that it will start developing 5nm MRAM technologies, to complement its 22-nm (in production), 16-nm (ready for customers verification) and ...
  77. [77]
    MRAM Evolves In Multiple Directions - Semiconductor Engineering
    Mar 11, 2021 · The main benefit of SOT MRAM technology is that the write current doesn't go through the MTJ – it passes parallel to the layers. That means that ...
  78. [78]
    Understanding Phase-Change Memory Alloys from a Chemical ...
    Sep 1, 2015 · Phase-change memories (PCM) are associated with reversible ultra-fast low-energy crystal-to-amorphous switching in GeTe-based alloys.
  79. [79]
    Phase-Change Memory from Molecular Tellurides | ACS Nano
    Phase-change memory (PCM) is an emerging memory technology based on the resistance contrast between the crystalline and amorphous states of a material.
  80. [80]
    An overview of phase-change memory device physics - IOPscience
    Mar 26, 2020 · In this article, we provide an overview of the current understanding of the main PCM device physics that underlie the read and write operations.
  81. [81]
    [PDF] Phase change materials and phase change memory - Eric Pop
    The current pulse heats the material by Joule heating, melts it, and enables very fast cooling (melt-quenching) such that the material solidifies in the ...
  82. [82]
    [PDF] Phase-Change Memory: Performance, Roles and Challenges
    Jul 16, 2019 · PCM provides a wide set of interesting features such as fast read and write access, excellent scalability potential, and high endurance recently ...
  83. [83]
    Electrical Phase-Change Memory: Fundamentals and State of the Art
    Aug 6, 2025 · The crystallization temperature (T c ) and melting temperature (T m ) of GST are reported to be 160 C and 600 C (Terao et al., 2009) (He et al., ...Missing: formula | Show results with:formula
  84. [84]
    The discovery of Ovshinsky switching and phase-change memory
    Jun 1, 2018 · An unusually broad set of experiences helped prepare Stanford Ovshinsky to recognize a phenomenon that forever changed solid-state physics.Missing: Samsung | Show results with:Samsung
  85. [85]
    Recent developments in phase‐change memory - Wiley Online Library
    Jun 22, 2022 · The first prototype was delivered from Intel and STMicroelectronics in 2008, offering 128 MB memory based on the 90 nm node [122]. This PCRAM ...Gesbte As A Typical Pcm... · Other Recent Pcm Materials · Pcm For Non-Von Neumann...<|separator|>
  86. [86]
    [PDF] A Survey of Phase Change Memory Systems - JCST
    Compared to DRAM, PCM (phase change memory) has better scalability, lower energy leakage, and non-volatility.
  87. [87]
    [PDF] Improving Write Operations in MLC Phase Change Memory
    However, despite many advantages, such as good scalability and low leakage, PCM suffers from ex- ceptionally slow write operations, which makes it challeng- ing ...
  88. [88]
    TaOx-based resistive switching memories: prospective and challenges
    The filamentary paths are formed under SET and ruptured under RESET. Electrochemical migration of oxygen ions and redox reaction near the metal/oxide interface ...
  89. [89]
    All HfOx-Resistive Switches with the Conducting Oxygen Vacancy ...
    Oct 5, 2025 · Resistive random-access memories (RRAM) based on binary metal oxides such as HfOx, (1−7) Y2O3, (8−11) TaOx, (12,13) and TiOx (14−16) are ...
  90. [90]
    Transition metal oxide based resistive random-access memory
    This paper provides an in-depth review of recent advancements in the field of RRAM, including the material used for fabrication and the methods to enhance the ...
  91. [91]
    Resistive Random Access Memory (RRAM) - PubMed Central - NIH
    Apr 22, 2020 · Resistive random access memory involves frequent transitions between a high resistance state (HRS) and low resistance state (LRS). Each ...
  92. [92]
    A review on device requirements of resistive random access ...
    Sep 6, 2023 · Among these requirements are a writing voltage of <3 V, a switching energy of <10 pJ and a time of <10 ns, a writing endurance of ∼1010 cycles, ...
  93. [93]
    Advances of embedded resistive random access memory in ...
    Mar 22, 2024 · Other emerging non-volatile memories such as PCM and MRAM are also being explored, but RRAM stands out due to its low dynamic power consumption, ...Missing: 1990s | Show results with:1990s
  94. [94]
    Panasonic Starts World's First Mass Production of ReRAM Mounted ...
    Jul 30, 2013 · Panasonic will start mass production in August 2013 of the ReRAM mounted microcomputers, contributing to longer operational times for security, portable and ...
  95. [95]
    NVM in Data Storage: A Post-Optane Future - ACM Digital Library
    Jul 22, 2025 · Companies like HP, Crossbar, Panasonic, Adesto Technologies, and Weebit Nano have been actively developing ReRAM technology for standalone ...
  96. [96]
    A Complete No-Brainer:ReRAM for Neuromorphic Computing | Weebit
    Jun 5, 2024 · ReRAM is a good match for neuromorphic applications due to its cost-efficiency, ultra-low power consumption, scaling advantage at 28nm and below, small ...
  97. [97]
    Resistive Random Access Memory (RRAM): an Overview of ...
    Apr 22, 2020 · Among all the emerging memory technology candidates, RRAM has significant advantages such as easy fabrication, simple metal-insulator-metal ( ...
  98. [98]
    Three-Dimensional Resistive Random-Access Memory Based on ...
    Apr 5, 2024 · However, RRAM typically suffers from various issues, such as a high-forming voltage and significant variation in switching behaviors. To ...
  99. [99]
    Understanding the resistive switching characteristics and ...
    Dec 18, 2012 · Conceptually, filament morphology can be altered to achieve higher resistance by reducing the filament cross sectional area, forming a gap in ...
  100. [100]
    [PDF] Modeling resistive switching materials and devices across scales
    In this model, set and reset switching kinetics are caused by the migration of oxygen vacancies. VO. The model schematizes the CF geometry as a VO top.
  101. [101]
    Resistive Switching Random-Access Memory (RRAM): Applications ...
    This work addresses the RRAM concept from materials, device, circuit, and application viewpoints, focusing on the physical device properties and the ...<|separator|>
  102. [102]
    Recent advancements and progress in development of ferroelectric ...
    The present study illustrates the fundamental concepts, operation, and construction of FeFETs and presents a methodology to determine suitable ferroelectric ...Missing: history | Show results with:history
  103. [103]
    FeFETs Bring Promise And Challenges - Semiconductor Engineering
    Feb 17, 2021 · FMC sees lower performance with pure HfO2, however. “We can use the hafnium oxide without anything, but then it limits our endurance,” said ...
  104. [104]
    High-performance ferroelectric field-effect transistors with ultra-thin ...
    Mar 27, 2024 · FeFET has high endurance of over 107 cycles and good retention characteristics estimated to be over ten years. Under different bending ...
  105. [105]
    An analytical interpretation of the memory window in ferroelectric ...
    Nov 29, 2023 · ... change in gate voltage ΔV is equivalent to the threshold voltage shift of the transistor. The basic idea for describing threshold voltage shift ...Missing: formula | Show results with:formula
  106. [106]
    What Are FeFETs? - Semiconductor Engineering
    Feb 16, 2017 · In comparison, the hafnium oxide based FeFET from FMC is a totally different memory cell in which the ferroelectric actually replaces the gate ...
  107. [107]
    Hafnium oxide-based ferroelectric field effect transistors
    Jul 7, 2025 · A FeFET is a single-transistor non-volatile memory device that integrates a ferroelectric material into the gate stack, allowing the ...
  108. [108]
    Magnetic Hard Disk - an overview | ScienceDirect Topics
    A magnetic hard disk is defined as a data storage device that relies on magnetism to store binary data on a circular platter made of magnetic material, ...
  109. [109]
    Giant Magnetoresistance - Magnet Academy - National MagLab
    MR read heads can detect much weaker magnetic fields than induction coil read heads, resulting in about four times more storage capacity. grunberg and fert.
  110. [110]
    Read/Write Head Technologies - Data Recovery Salon
    When writing, the current in the coil creates a polarized magnetic field in the gap between the poles of the core, which magnetizes the surface of the platter ...
  111. [111]
    vol.2 Magnetic head technology paved the way for smaller hard disk ...
    The basic components of a HDD are a magnetic disk spinning at high speed, and a swing arm that scans the disk for reading and writing data. Over the past 20 ...
  112. [112]
    RAMAC - IBM
    or simply RAMAC — was the first computer to use a random-access disk drive. The progenitor of all hard disk drives created since, it made it ...
  113. [113]
    Seagate launches 28 and 30 TB HAMR hard drives for edge AI and ...
    Jul 15, 2025 · HAMR (heat-assisted magnetic recording) uses a laser to temporarily heat a high coercivity bit area, reducing its resistance to magnetic state ...
  114. [114]
    Exos M | Seagate US
    Seagate's Exos M hard drive, boasting breakthrough 3TB per platter density, delivers extraordinary storage capacity and power efficiency.
  115. [115]
    Areal density optimizations for heat-assisted magnetic recording of ...
    Jun 14, 2016 · Heat-assisted magnetic recording (HAMR) is hoped to be the future recording technique for high-density storage devices.
  116. [116]
    LTO-9: LTO Generation 9 Technology | Ultrium LTO - LTO.org
    This means an LTO-9 tape can deliver the same capacity with only 1/85th of the areal density of a similar capacity disk. Simply, tape has the headroom to expand ...
  117. [117]
    LTO-9 | Quantum
    LTO-9 Capacity. Quantum LTO-9 offers the latest in LTO technology and delivers increased tape cartridge capacity with up to 18 TB (45 TB* compressed).
  118. [118]
    What difference does areal density make? | Seagate US
    Drives with higher areal density deliver 2.6× better power efficiency per terabyte than current 10TB drives.Missing: time 5-10 ms
  119. [119]
    Optical data storage - Physics World
    Oct 1, 1998 · This is achieved by recording the data using a laser that creates pits in a disc substrate covered with a carefully selected film of organic ...
  120. [120]
    [PDF] Optical Disks for Image Storage and Distribution - IS&T | Library
    On a CD, as on a video disk, the information is stored as “pits” in a substrate. The 120-mm-diameter, 1.2-mm- thick CD substrate is made of polycarbonate, and ...
  121. [121]
    When Were CDs Invented, and How Do They Work? | HowStuffWorks
    Mar 6, 2024 · Developed jointly by Philips and Sony, the CD format was finalized in 1982 and famously documented in the Red Book standard.
  122. [122]
    5.6.1 Introduction
    5.6.1.1 Since their introduction in 1982 replicated optical disc media have become the dominant technology for distribution of published audio recordings.
  123. [123]
    Super Multi Blu-ray Writer - Hitachi-LG Data Storage
    Ultimate Capacity. Three Layer BDXL-R/BDXL RE discs offer 100GB capacity while quad layer BDXL-R discs boast 128GB storage capacity on a single disc. It is ...
  124. [124]
    How do rewriteable CDs work? - Scientific American
    Aug 26, 2002 · Compact Disc Rewriteable (CD-RW) is a fully rewriteable media, meaning that any spot on a CD-RW disc can be rewritten up to 1,000 times (based ...Missing: cycles endurance<|separator|>
  125. [125]
    Optical Storage | Research Starters - EBSCO
    Most optical storage is in the form of optical disks, which are flat and circular. They contain binary data in the form of microscopic pits, which are non- ...Missing: lands substrate
  126. [126]
    The range of access times of optical disks is generally ______
    Feb 20, 2022 · Easy explanation: The access times of optical disks are typically in the range of 100 to 300 milliseconds and that of hard disks are in the ...<|separator|>
  127. [127]
    Microscope Resolution: Concepts, Factors and Calculation
    In order to increase the resolution, d = λ/(2NA), the specimen must be viewed using either a shorter wavelength (λ) of light or through an imaging medium with ...
  128. [128]
    Digital Storage And Memory Projections For 2025, Part 2 - Forbes
    Dec 11, 2024 · Projections of new optical storage archiving systems promise 1PB optical cartridges by the 2030's, compared to the possible 576TB cartridge ...
  129. [129]
    Why aren't optical disks the top choice for archive storage?
    Oct 26, 2021 · Optical storage can last a century, give faster access to data, and is backward compatible with earlier optical technology, yet disk drives ...
  130. [130]
    Advantages and disadvantages of Optical Disks - GeeksforGeeks
    Jul 23, 2025 · The optical discs remain a popular type of storage medium because they are relatively cheap, relatively durable and easy to transport.
  131. [131]
    Organic Non‐Volatile Memory Based on Pentacene Field‐Effect ...
    Nov 10, 2006 · An organic field-effect transistor (OFET) memory device based on pentacene is fabricated using an additional poly(α-methyl styrene) gate ...Missing: principle | Show results with:principle
  132. [132]
    Electrical memory devices based on inorganic/organic ... - Nature
    Jun 8, 2012 · This review article describes investigations of and developments in nonvolatile memory devices based on hybrid inorganic/organic nanocomposites over the past 5 ...Device Structure And... · Electrical Characteristics... · Switching And Carrier...
  133. [133]
    Pulse-modulated multilevel data storage in an organic ferroelectric ...
    Apr 15, 2016 · We demonstrate multilevel data storage in organic ferroelectric resistive memory diodes consisting of a phase-separated blend of P(VDF-TrFE) ...
  134. [134]
    Nonvolatile Memory Elements Based on Organic Materials - Scott
    May 10, 2007 · This Review summarizes the materials that have been used in switching devices, and describes the variety of device behavior observed in their ...
  135. [135]
    Organic Memory Devices: A Leap Forward in Wearable Computing
    Sep 17, 2025 · Brain-inspired electronics with organic memristors could offer a functionally promising and cost- effective platform, according to Venkatesan.
  136. [136]
    Printed and flexible organic and inorganic memristor devices for non ...
    Sep 28, 2023 · In this article we review progress in the field of printed and flexible memristor devices and discuss their potential across a wide range of applications.
  137. [137]
    Resolving electron and hole transport properties in semiconductor ...
    Jan 5, 2024 · We focused on the mobility and lifetime of charge carriers as shown in Fig. 5. The observed ranges of mobility spanned from 10–3 cm2V–1s–1 in ...
  138. [138]
    Micron Ends 3D XPoint Memory - Forbes
    Mar 16, 2021 · Micron announced that it was ending its development of the 3D XPoint technology that it co-developed with Intel.Missing: 2022 selector
  139. [139]
    Intel announces the discontinuation of Optane, 3D Xpoint finally ...
    Oct 17, 2022 · Intel has just issued a product discontinuation notice for Intel Optane 100 DIMMs. These are the DIMMs launched in 2019 with the ...<|separator|>
  140. [140]
    Intel 3D XPoint Memory Die Removed from Intel Optane™ PCM ...
    May 18, 2017 · TechInsights recently acquired and tore down an Intel OptaneTM M.2 80mm 16GB PCIe 3.0 and discovered a 3D X-Point memory die in the package.
  141. [141]
    Single-molecule non-volatile memories: an overview and future ...
    This article provides an overview of the research progress on non-volatile memories associated with functional centers of single molecules.Missing: redox | Show results with:redox
  142. [142]
    Assessing Design Space for the Device-Circuit Codesign of ...
    Jan 13, 2025 · Compute-in-memory (CIM) architectures further improve energy efficiency by fusing the computation operations with AI model storage. Electronic ...
  143. [143]
    Design of 1T2R ReRAM array for in memory element-wise ...
    Aug 6, 2025 · This study considers an in-memory element-wise multiplication using a One Transistor two Resistor (1T2R) ReRAM array. The adder-shifter module ...
  144. [144]
    Media Advisory - SMART Modular to Showcase CXL® Persistent ...
    Jul 30, 2025 · Newark, Calif. – August 5, 2025 Live demo will show how CXL persistent memory can be used as a high-performance checkpoint solution versus ...
  145. [145]
    CXL is Finally Coming in 2025 - ServeTheHome
    Dec 19, 2024 · Still, in 2025, expect to see more CXL server designs for those who need more memory and memory bandwidth in general purpose compute. We will ...Missing: persistent trends
  146. [146]
    Samsung on track to start 14 nm eMRAM production by the end of ...
    Jun 1, 2024 · The company now says that it has finalized the development of its 14 nm eMRAM process, and will be read to mass produce it by the end of 2024.Missing: LPDDR mobile
  147. [147]
    Persistent Memory Market Analysis (2035) - WiseGuy Reports
    The Global Persistent Memory Market is projected to grow at a CAGR of 10.1% from 2025 to 2035, driven by the increasing demand for high-performance computing ...
  148. [148]
    9100 PRO | Internal SSD | Samsung Semiconductor Global
    Experience exceptional speed with the Samsung 9100 PRO SSD, featuring PCIe 5.0 and sequential read/write speeds up to 14800/13400 MB/s.
  149. [149]
    Western Digital Black SN8100 gets an 8TB capacity boost — fastest ...
    Oct 21, 2025 · This version of the drive is capable of pushing a positively massive 14,900 MB/s for sequential reads and 13,200 MB/s for sequential writes.
  150. [150]
    DataTraveler Max USB 3.2 Gen 2 Series Flash Drive
    Free delivery 30-day returnsAvailable in USB-C ® and USB-A options, it offers high-capacity storage up to 1TB in a compact, portable design–perfect for HD photos, 4K/8K videos, music and ...
  151. [151]
    Best Flash Drives 2025: Fast, Spacious, Pocketable USB Storage
    Jul 30, 2025 · We've tested dozens of recent flash drives from leading brands and lesser-known alternatives and listed the best USB flash drives below, to help narrow down ...Best Flash Drives You Can... · 2. Sk Hynix Tube T31 (1tb) · 3. Adata Sc750
  152. [152]
    Nand Flash Memory Market Size, Trends & Forecast, 2025-2032
    Jun 18, 2025 · The NAND flash memory market is estimated to be valued at USD 73.41 Bn in 2025 and is expected to reach USD 109.01 Bn by 2032, exhibiting a compound annual ...
  153. [153]
    3Q25 NAND Flash Contract Prices Projected to Rise 5–10%
    Jul 9, 2025 · 3Q25 NAND Flash Contract Prices Projected to Rise 5–10%; Weak Smartphone Demand Limits eMMC and UFS Growth, Says TrendForce.<|separator|>
  154. [154]
    Tape is Primed for the Rise of Secondary Storage - Backupworks.com
    According to the report, tape produces 97% less CO2e compared to an all-HDD storage strategy! With insatiable secondary storage demand ahead, expect tape to ...
  155. [155]
    How Does Hardware-Based SSD Encryption Work? Software vs ...
    Learn how hardware-based SSD encryption uses AES 256-bit and TCG Opal 2.0 for secure, efficient, and tamper-proof storage.
  156. [156]
    Emerging Memory Takes the Embedded Route - IDTechEx
    May 29, 2025 · The report analyzes key technologies such as MRAM, ReRAM, FeRAM, and PCM, as well as the leading players commercializing them and their impact ...Missing: FRAM history milestones 1990s 2000s 2010s
  157. [157]
    Samsung's revived Z-NAND targets 15x performance increase over ...
    Aug 8, 2025 · Samsung is reviving its Z-NAND memory technology with a next-generation version that could be up to 15x faster than today's PCIe 4.0 or 5.0 ...
  158. [158]
    Solidigm Presents Optane Replacement - Tech Field Day
    Solidigm is positioning this product in conjunction with their Cloud Storage Acceleration Layer as a viable replacement for Optane. Read More. Solidigm Presents ...Missing: concepts | Show results with:concepts<|separator|>
  159. [159]
    Persistent Memory vs RAM in 2025: CXL & NVDIMM-P Guide
    May 29, 2025 · In 2025, the two main PMEM options are NVDIMM-P (plugs into DDR5 slots) and CXL Type-3 (fabric-attached via PCIe).Missing: trends | Show results with:trends
  160. [160]
    CXL Non-Volatile Memory Module - Viking Technology
    Explore Viking Technology's CXL Non-Volatile Memory Module, a standard Type 3 memory expander Rev 2.0 with persistent memory capability.
  161. [161]
    Introduction To CXL 2.0 Memory - Lenovo Press
    Feb 24, 2025 · CXL offers the promise of non-volatile memory which is not attached to the memory bus. The advantage is that it does not limit the amount of ...Missing: NVDIMM | Show results with:NVDIMM
  162. [162]
    NV-CMM-E3S | CXL®-Based Solutions from SMART Modular
    SMART's NV-CMM-E3S non-volatile memory module with built-in backup power source utilizes the Compute Express Link (CXL)® 2.0 standard in an E3.S 2T form factor.Missing: NVDIMM | Show results with:NVDIMM
  163. [163]
    [PDF] A Study of Application Performance with Non-Volatile Main Memory
    We find that although NVMM is projected to have higher latency and lower bandwidth than DRAM, these difference have only a modest impact on application ...
  164. [164]
    Analyzing and modeling the impact of memory latency and ...
    NVM (Non-Volatile Memory) that gains attention as an alternative memory technology these days has different memory characteristics in terms of latency and ...
  165. [165]
  166. [166]
  167. [167]
    2023 IRDS Mass Data Storage
    Although this process does scale up in total chip density each generation, it is less cost effective than the density increase achieved. Page 17. 9. THE ...
  168. [168]
    NXP and TSMC to Deliver Industry's First Automotive 16 nm FinFET ...
    May 16, 2023 · TSMC's 16FinFET embedded MRAM technology exceeds the rigorous requirements of automotive applications with its one-million cycle endurance, ...
  169. [169]
    Memory, mram, page 1-Research-Taiwan Semiconductor ...
    We demonstrate the reliability and magnetic immunity of STT-MRAM embedded in 16nm FinFET CMOS process. The technology supports endurance cycles up to 105 for ...
  170. [170]
    [PDF] Performance and Reliability in Demanding Automotive Applications
    • AEC Q-100 Qualified Grade 1 (-40/125°C) and Grade 3 (-40/85°C) available. Protect Valuable Data and Reduce Costs. MRAM memory is the best solution to ...
  171. [171]
    Infineon Puts FRAM in Space - EE Times
    Jun 17, 2022 · Infineon claims its QML–V qualified FRAM is the space industry's first radiation–hardened (rad–hard) FRAM.<|separator|>
  172. [172]
    Memories (SRAM, MRAM) - Honeywell Aerospace
    Our radiation-hardened memories provide aerospace and military systems highly reliable, solutions for intense radiation environments. Read more!
  173. [173]
    MRAM for Radiation-hard Markets - Everspin
    Magnetoresistive RAM (MRAM) is a non-volatile memory that can protect data from loss of power and perform appropriately in space.
  174. [174]
    Eco-Friendly Biomemristive Nonvolatile Memory - ACS Publications
    Jul 8, 2024 · Here, natural organically grown orange peel was used to extract biocompatible pectin to design a resistive switching-based memory device.
  175. [175]
    Bioresorbable Resistive Switching Device Based on Organic ...
    Jan 23, 2024 · Biocompatible and ... organic/inorganic structures are reported, showing nonvolatile bipolar resistive switching memory behavior.
  176. [176]
    Low Power Non-volatile Memory Technology for Edge AI Application
    Deep neural network (DNN) inference for edge AI requires low power operation. We demonstrated it by implementing massively parallel matrix-vector ...
  177. [177]
    [PDF] A 28 nm AI microcontroller with tightly coupled zero-standby ... - arXiv
    The AI model can be stored and updated in an embedded Non-Volatile Memory (eNVM) during the device's lifetime without consuming standby power during the idle ...Missing: 1μW | Show results with:1μW