Search

Superlattices made of alternating blocks of the phase change compound Sb 2 Te 3 and of TiTe 2 confining layers have been recently proposed for applications in neuromorphic devices. The Sb 2 Te 3 /TiTe 2 heterostructure allows for a better control of multiple intermediate resistance states and for a lower drift with time of the electrical resistance of the amorphous phase. However, Sb 2 Te 3 suffers from a low data retention due to a low crystallization temperature T x . Substituting Sb 2 Te 3 with a phase change compound with a higher T x , such as GeTe, seems an interesting option in this respect. Nanoconfinement might, however, alters the crystallization kinetics with respect to the bulk. In this work, we investigated the crystallization process of GeTe nanoconfined in geometries mimicking GeTe/TiTe 2 superlattices by means of molecular dynamics simulations with a machine learning potential. The simulations reveal that nanoconfinement induces a mild reduction in the crystal growth velocities which would not hinder the application of GeTe/TiTe 2 heterostructures in neuromorphic devices with superior data retention. [ABSTRACT FROM AUTHOR]

The doping with Cu in the Sb-Se material system can effectively reduce power consumption and improve data retention ability of the PCRAM (Phase-Change Random-Access Memory) material. However, no thermodynamic study of the Cu-Sb-Se system has been reported. Therefore, in this work the Cu-Sb-Se system was modeled using the CALPHAD (CALculation of PHAse Diagrams) method. The excess Gibbs energies of solution phases, including liquid and FCC were expressed by the Redlich-Kister polynomial. Two-sublattice model (Cu,Se)2Se was used to describe the solid solution of binary intermetallic compounds, i.e. α-Cu2Se and β-Cu2Se in Cu-Sb-Se ternary system. The compounds, e.g. η(Cu2Sb), β(Cu3Sb), Sb2Se3, α-CuSe, β-CuSe, Pe(Cu3SbSe4) and C(CuSb3Se5) were described as stoichiometric compounds. The temperature of the eutectic reaction γ(Cu17Sb3) = (Cu) + δ(Cu4Sb) in the Cu-Sb system and the temperature of the invariant reaction L1 = L2 + (Sb) in the Sb-Se system were adjusted. A set of self-consistent thermodynamic parameters for the Cu-Sb-Se system was obtained, and the 623, 673, 723, 773, 873 and 973 K isothermal sections, and the liquidus projection and invariant reactions for this ternary system have been calculated. [ABSTRACT FROM AUTHOR]

Phase Change Memory (PCM) is an emerging technology exploiting the rapid and reversible phase transition of certain chalcogenides to realize nanoscale memory elements. PCM devices are being explored as non-volatile storage-class memory and as computing elements for in-memory and neuromorphic computing. It is well-known that PCM exhibits several characteristics of a memristive device. In this work, based on the essential physical attributes of PCM devices, we exploit the concept of Dynamic Route Map (DRM) to capture the complex physics underlying these devices to describe them as memristive devices defined by a state—dependent Ohm's law. The efficacy of the DRM has been proven by comparing numerical results with experimental data obtained on PCM devices. [ABSTRACT FROM AUTHOR]

Phase Change Memory (PCM) is an emerging technology exploiting the rapid and reversible phase transition of certain chalcogenides to realize nanoscale memory elements. PCM devices are being explored as non-volatile storage-class memory and as computing elements for in-memory and neuromorphic computing. It is well-known that PCM exhibits several characteristics of a memristive device. In this work, based on the essential physical attributes of PCM devices, we exploit the concept of Dynamic Route Map (DRM) to capture the complex physics underlying these devices to describe them as memristive devices defined by a state—dependent Ohm's law. The efficacy of the DRM has been proven by comparing numerical results with experimental data obtained on PCM devices. [ABSTRACT FROM AUTHOR]

Te-based alloys are widely used for Phase Change Random Access Memory (PCRAM). The investigation of the Te-based phase diagram gives significant benefit to the development of Te-based PCRAM materials. Thermodynamic modelling of the Te-X (X = Cu, Ga, Li, Sr) systems has been carried out using the CALculation of PHAse Diagrams (CALPHAD) method based on the enthalpies of formation calculated by first-principles calculations in the present work and available experimental data in the literature. In the Te-Cu, Te-Ga and Te-Li systems, the liquid phases were separately described by the associated solution model with Cu2Te, Ga2Te3 and Li2Te associate species. A set of self-consistent thermodynamic parameters for the Te-X (X = Cu, Ga, Li, Sr) systems was obtained. Comparison between the calculated results and available experimental data indicates that most experimental information can be satisfactorily accounted for by modelling. [ABSTRACT FROM AUTHOR]

This research is to design a history table-based linear analysis method for a DRAM-PCM (phase change memory) hybrid memory system supporting irregular and complex memory accessing patterns in graph processing applications. This linear analysis method can analyze complex memory access patterns in real time through our history table structure, classifying address values requested from the LLC (last-level cache) as the prefetch candidates. Also, the proposed method is to support a prefetching operation from the storage onto the hybrid main memory, causing significant latency losses. Specifically, we proposed a new memory pattern analysis method, which is called a linear analysis based on a history table, to detect any linear memory access patterns from workloads showing irregular accessing characteristics. For the DRAM-PCM hybrid memory, dynamic data that are frequently accessed are stored in the DRAM. Thus, our linear analysis not only categorizes memory access patterns in the form of line patterns but also separates dynamic and static data so that the hybrid memory systems can be managed adaptively and efficiently. Our experiments show that our model can improve performance and energy consumption by 60% and 30%, respectively, compared to those of the traditional DRAM-only memory model. Moreover, our approach can enhance the entire system performance and energy consumption by 11.3%, 6.8%, compared to streamlined prefetcher-based models. [ABSTRACT FROM AUTHOR]

A variety of resistive memories have been proposed in recent years. Among these emerging technologies, phase change memory (PCM) has received the most research attentions since it has the advantages of high scalability, non-volatility, fast access, strong data retention, low cost, and low power consumption. It is also considered as the most promising alternative of DRAM. In order to conquer the inevitable endurance problem of PCM cells which causes serious reliability and yield threats, hard repair and ECC (Error correction code) techniques are widely adopted. However, since soft errors are not a main threat for PCM, incorporating ECC for each data word is not a cost-effective technique since a lot of memory space is required for storing the check bits. In this paper, the progressive ECC techniques including the local progressive ECC (LPE) technique and the global progressive ECC (GPE) technique are proposed to solve this dilemma. The key innovation is to equip ECC for a data word when its first faulty cell is detected. In other words, we only equip fault detection code for data words such that the original code rate can be increased significantly. An ECC DRAM and an ECC CAM are used for storing check bits and accessing purposes, respectively. Hardware architectures for implementing the proposed GPE and LPE techniques are also provided. A simulator is developed for evaluating repair rate, reliability, yield, and hardware overhead. According to experimental results, the degradation of repair rate and reliability are almost negligible. However, the hardware overhead is at least 80% lower than the original ECC technique while maintaining the original reliability and yield levels. [ABSTRACT FROM AUTHOR]

Phase change materials exhibit threshold switching (TS) that establishes electrical conduction through amorphous material followed by Joule heating leading to its crystallization (set). However, achieving picosecond TS is one of the key challenges for realizing non-volatile memory operations closer to the speed of computing. Here, we present a trajectory map for enabling picosecond TS on the basis of exhaustive experimental results of voltage-dependent transient characteristics of Ge2Sb2Te5 phase-change memory (PCM) devices. We demonstrate strikingly faster switching, revealing an extraordinarily low delay time of less than 50 ps for an over-voltage equal to twice the threshold voltage. Moreover, a constant device current during the delay time validates the electronic nature of TS. This trajectory map will be useful for designing PCM device with SRAM-like speed. [ABSTRACT FROM AUTHOR]

Phase-change memory utilizing amorphous-to-crystalline phase-change processes for reset-to-set operation as a nonvolatile memory has been recently commercialized as a storage class memory. Unfortunately, designing new phase-change materials (PCMs) with low phase-change energy and sufficient thermal stability is difficult because phase-change energy and thermal stability decrease simultaneously as the amorphous phase destabilizes. This issue arising from the trade-off relationship between stability and energy consumption can be solved by reducing the entropic loss of phase-change energy as apparent in crystalline-to-crystalline phase-change process of a GeTe/Sb2Te3 superlattice structure. A paradigm shift in atomic crystallography has been recently produced using a quasi-crystal, which is a new type of atomic ordering symmetry without any linear translational symmetry. This paper introduces a novel class of PCMs based on a quasicrystalline-to-approximant crystalline phase-change process, whose phase-change energy and thermal stability are simultaneously enhanced compared to those of the GeTe/Sb2Te3 superlattice structure. This report includes a new concept that reduces entropic loss using a quasicrystalline state and takes the first step in the development of new PCMs with significantly low phase-change energy and considerably high thermal stability. [ABSTRACT FROM AUTHOR]

Phase change memory (PCM) is being actively explored for in-memory computing and neuromorphic systems. The ability of a PCM device to store a continuum of resistance values can be exploited to realize arithmetic operations such as matrix-vector multiplications or to realize the synaptic efficacy in neural networks. However, the resistance variations arising from structural relaxation, 1/f noise, and changes in ambient temperature pose a key challenge. The recently proposed projected PCM concept helps to mitigate these resistance variations by decoupling the physical mechanism of resistance storage from the information-retrieval process. Even though the device concept has been proven successfully, a comprehensive understanding of the device behavior is still lacking. Here, we develop a device model that captures two key attributes, namely, resistance drift and the state dependence of resistance. The former refers to the temporal evolution of resistance, while the latter refers to the dependence of the device resistance on the phase configuration of the phase change material. The study provides significant insights into the role of interfacial resistance in these devices. The model is experimentally validated on projected PCM devices based on antimony and a metal nitride fabricated in a lateral device geometry and is also used to provide guidelines for material selection and device engineering. [ABSTRACT FROM AUTHOR]

Spiking neural networks (SNN) are computational models inspired by the brain's ability to naturally encode and process information in the time domain. The added temporal dimension is believed to render them more computationally efficient than the conventional artificial neural networks, though their full computational capabilities are yet to be explored. Recently, in-memory computing architectures based on non-volatile memory crossbar arrays have shown great promise to implement parallel computations in artificial and spiking neural networks. In this work, we evaluate the feasibility to realize high-performance event-driven in-situ supervised learning systems using nanoscale and stochastic analog memory synapses. For the first time, the potential of analog memory synapses to generate precisely timed spikes in SNNs is experimentally demonstrated. The experiment targets applications which directly integrates spike encoded signals generated from bio-mimetic sensors with in-memory computing based learning systems to generate precisely timed control signal spikes for neuromorphic actuators. More than 170,000 phase-change memory (PCM) based synapses from our prototype chip were trained based on an event-driven learning rule, to generate spike patterns with more than 85% of the spikes within a 25 ms tolerance interval in a 1250 ms long spike pattern. We observe that the accuracy is mainly limited by the imprecision related to device programming and temporal drift of conductance values. We show that an array level scaling scheme can significantly improve the retention of the trained SNN states in the presence of conductance drift in the PCM. Combining the computational potential of supervised SNNs with the parallel compute power of in-memory computing, this work paves the way for next-generation of efficient brain-inspired systems. [ABSTRACT FROM AUTHOR]

Phase change memory (PCM) offers remarkable features such as high-speed and non-volatility for universal memory. Yet, simultaneously achieving better thermal stability and fast switching remains a key challenge. Thus, exploring novel materials with improved characteristics is of utmost importance. We report here, a unique property-portfolio of high thermal stability and picosecond threshold switching characteristics in In3SbTe2 (IST) PCM devices. Our experimental findings reveal an improved thermal stability of amorphous IST compared to most other phase change materials. Furthermore, voltage dependent threshold switching and current-voltage characteristics corroborate an extremely fast, yet low electric field threshold switching operation within an exceptionally small delay time of less than 50 picoseconds. The combination of low electric field and high speed switching with improved thermal stability of IST makes the material attractive for next-generation high-speed, non-volatile memory applications. [ABSTRACT FROM AUTHOR]

Phase change memory (PCM) is one of promising technology to replace DRAM with its attractive features such as zero leakage power and high scalability. In PCM, a SET operation needs much more time than a RESET operation. A typical write request concurrently writes 64 bytes to a PCM memory line. Therefore, write latency is mainly determined by SET operations. Previously, PreSET has been proposed to improve PCM performance by exploiting asymmetry in write time. A PreSET operation pro-actively SETs all the bits in the memory line before a dirty cache line is written to PCM. Later, when a write request is processed, only RESET operations are actually performed. Consequently, PreSET reduces write latency and improves system performance. However, such PreSET operations are conducted only at a very coarse-grained level, which reduces the endurance of PCM. Through empirical study, we observe that in most applications the number of dirty words in a dirty line is actually quite limited. If we only SET those dirty words, instead of the whole cache line, we would significantly extend the lifetime of PCM while still achieving desirable performance. Inspired by this observation, we propose a scheme called Partial-PreSET which balances performance and endurance of PCM. The core idea of this scheme is to SET those dirty bits of a cache line in a fine-grained fashion. Our experiments show that the proposed Partial-PreSET scheme significantly improves the average lifetime of PCM system, up to 2.79X, while incurring only 2% system performance loss, compared with the state-of-the-art scheme (i.e., PreSET). [ABSTRACT FROM AUTHOR]

The new emergence of data-intensive applications raises a huge requirement on the capacity of memory. However, an obvious increase of memory makes the corresponding energy consumption unacceptable in practice. To address this question, any effort only to reduce the energy consumption of dynamic random access memory (DRAM) is ineffective. Recently, combining DRAM and phase change memory (PCM) to construct a hybrid main memory is recognized as a promising solution to distinctly reduce the energy consumption. In this paper, we propose a new page-level energy management strategy to optimize the energy consumption of the hybrid main memory. The proposed strategy records pages' local and global access information by a new data structure, and then classifies pages by the access history, at last adaptively places PCM or DRAM pages according to the memory characteristics and remaps the migrated pages. Our experimental results show that our strategy can achieve 9.4 % of energy saving and 9.6 % of performance improvement at most compared with APG and PDRAM, which were proposed respectively by conferences RACS'12 and DAC'09. [ABSTRACT FROM AUTHOR]

We propose a method for evaluating kinetic parameters for the crystallization of thin films of phase change materials. Its basic principle is to jointly use model-free and model isoconversional methods in analyzing differential scanning calorimetry results. Using this method, we have identified the reaction model and evaluated the activation energy for crystallization and pre-exponential factor as a function of the degree of conversion for GeSbTe-based thin films. [ABSTRACT FROM AUTHOR]

Phase memory in a three-level system that is associated with the correspondence of isochromates of inhomogeneously broadened lines excited by lasers at various resonant frequencies with a common energy level in different time intervals is studied. It is shown that external spatially inhomogeneous electric fields can control such phase memory and could be used to determine the optimum conditions for forming a stimulated photon echo in a threelevel system. [ABSTRACT FROM AUTHOR]

In this letter, a phase change random access memory (PCRAM) chip based on TiSbTe alloy material was fabricated in a 40-nm 4-metal level complementary metal-oxide semiconductor (CMOS) technology. The phase change resistor was then integrated after CMOS logic fabrication. The PCRAM was successfully embedded without changing any logic device and process, in which 1.1 V negative-channel metal-oxide semiconductor device was used as the memory cell selector. The currents and the time of SET and RESET operations were found to be 0.2 and 0.5 mA, 100 and 10 ns, respectively. The high speed performance of this chip may highlight the design advantages in many embedded applications. [ABSTRACT FROM AUTHOR]

Reconfigurable models were shown to be very powerful in solving many problems faster than non-reconfigurable models. WECPAR $$W(M,N,k)$$ is an $$M\times N$$ reconfigurable model that has point-to-point reconfigurable interconnection with $$k$$ wires between neighboring processors. This paper studies several aspects of WECPAR. We first consider solving the list ranking problem on WECPAR. Some of the results obtained show that, ranking one element in a list of $$N$$ elements can be solved on $$W(N,N,N)$$ WECPAR in $$O(1)$$ time. Also, on $$W(N,N,k)$$ , ranking a list $$L(N)$$ of $$N$$ elements can be done in $$O((\log N)(\lceil {\log _{k+1} N} \rceil ))$$ time. Then, we assess the relative computational power of WECPAR and transfer a large body of algorithms to work directly on WECPAR. We introduce several simulation algorithms between WECPAR and well-known models such as PRAM and RMBM. Simulation algorithms show that a PRIORITY CRCW PRAM $$P(N,S)$$ of $$N$$ processors and $$S$$ shared memory locations can be simulated on $$W(S,N,k)$$ WECPAR in $$O(\lceil {\log _{k+1} N}\rceil +\lceil {\log _{k+1} S}\rceil )$$ time. Also, we show that a PRIORITY CRCW basic RMBM( $$P,B)$$ , of $$P$$ processors and $$B$$ buses can be simulated on $$W(B, P+B, k)$$ WECPAR in $$O(\lceil {\log _{k+1} (P+B)}\rceil )$$ time. This directly migrate a large number of algorithms to work on WECPAR with the simulation overhead. [ABSTRACT FROM AUTHOR]

Interfacial phase change memory (iPCM), that has a structure of a superlattice made of alternating atomically thin GeTe and Sb2Te3 layers, has recently attracted attention not only due to its superior performance compared to the alloy of the same average composition in terms of energy consumption but also due to its strong response to an external magnetic field (giant magnetoresistance) that has been speculated to arise from switching between topological insulator (RESET) and normal insulator (SET) phases. Here we report magneto-optical Kerr rotation loops in the visible range, that have mirror symmetric resonances with respect to the magnetic field polarity at temperatures above 380 K when the material is in the SET phase that has Kramers-pairs in spin-split bands. We further found that this threshold temperature may be controlled if the sample was cooled in a magnetic field. The observed results open new possibilities for use of iPCM beyond phase-change memory applications. [ABSTRACT FROM AUTHOR]

The bond constraint theory (BCT) dealing with the rigidity caused by bond constraints and the long-range potential fluctuations (LRPF) arising from the defects and heterogeneities in the disordered semiconductors are important for understanding the atomic and electronic properties of amorphous semiconductors. Here, they are applied to the already commercialized GeSbTe (GST) chalcogenide glasses used in the rewritable phase change memory (PCM) devices. The main concern at present is to improve their ability to withstand a large number of phase change cycles, by choosing the right composition. The two considerations (BCT and LRPF) are briefly described and tested on the most commonly used GeSbTe and the nearby compositions. While these considerations provide significant insight into their atomic and electronic structures, the ansatz linking them with the durability of the PCM devices need to be justified by more work. [ABSTRACT FROM AUTHOR]

Electrical and switching property of amorphous defect chalcopyrite ZnGa2Te4 thin films prepared by thermal evaporation technique has been studied. The elemental chemical compositions of the prepared bulk as well as the as-deposited film were determined by means of energy dispersive X-ray spectrometry. X-ray diffraction pattern revealed that the powder compound is polycrystalline and the as-deposited and the annealed films at t ≤ 548 K have the amorphous phase, while that the annealed at t ≥ 573 K are polycrystalline with a single phase of a defect chalcopyrite structure similar to that of the synthesized material. The great advantage of this material is the capability to appear in two different phases, the amorphous and the crystalline phases, with rather different electrical properties. Both dynamic and static I-V characteristics and the switching phenomenon at 601 nm are investigated. The threshold switching mechanism was explained by a thermal model of switching, i.e., joule heating with an electrically conducting channel. ZnGa2Te4 is good candidate in phase change memory device. [ABSTRACT FROM AUTHOR]

GeSbTe (GST) is a technologically important phase-change material for data storage, where the fast reversible phase transition between crystalline and amorphous states is used for recording information. The encoding is achieved by the large contrast in physical properties between the two states. Ge vacancies (V) and Sb antisite atoms (Sb) are primary point defects in crystalline GST. The effect of V and Sb on the atomic arrangements in amorphous GST is unknown, which, however, has significant effects on the performance of GST. In this work, by means of ab initio molecular dynamics calculations, the atomic arrangements in amorphous ideal and defective GST have been investigated. The results show that the amorphous structure of GST with Sb ( St-GST) or with V ( V-GST) has the same cubic framework and close chemical ordering to ideal GST, and hence similar fast reversible phase transition behavior is expected in the defective phases. Furthermore, the presence of Sb or V in the crystalline phase will result in much more Ge atoms in a tetrahedral geometry as well as in a fourfold octahedral environment in the amorphous state. Especially in V-GST, around 36 % Ge atoms occupy a fourfold octahedral geometry. As fourfold octahedral Ge atoms can enhance the large contrast in physical properties between the crystalline and amorphous phases, introducing Ge vacancies would be an efficient way to improve the performance of GST phase-change materials. [ABSTRACT FROM AUTHOR]

Modern processors such as Tilera's Tile64, Intel's Nehalem, and AMD's Opteron are migrating memory controllers (MCs) on-chip, while maintaining a large, flat memory address space. This trend to utilize multiple MCs will likely continue and a core or socket will consequently need to route memory requests to the appropriate MC via an inter- or intra-socket interconnect fabric similar to AMD's HyperTransport, or Intel's Quick-Path Interconnect. Such systems are therefore subject to non-uniform memory access (NUMA) latencies because of the time spent traveling to remote MCs. Each MC will act as the gateway to a particular region of the physical memory. Data placement will therefore become increasingly critical in minimizing memory access latencies. Increased competition for memory resources will also increase the memory access latency variation in future systems. Proper allocation of workload data to the appropriate MC will be important in decreasing the variation and average latency when servicing memory requests. The allocation strategy will need to be aware of queuing delays, on-chip latencies, and row-buffer hit-rates for each MC. In this paper, we propose dynamic mechanisms that take these factors into account when placing data in appropriate slices of physical memory. We introduce adaptive first-touch page placement, and dynamic page-migration mechanisms to reduce DRAM access delays for multi-MC systems. We also introduce policies that can handle data placement in memory systems that have regions with heterogeneous properties. The proposed policies yield average performance improvements of 6.5% for adaptive first-touch page-placement, and 8.9% for a dynamic page-migration policy for a system with homogeneous DRAM DIMMs. We also show improvements in systems that contain DIMMs with different performance characteristics. [ABSTRACT FROM AUTHOR]

Three-dimensional stackable memory frames involving the integration of two-terminal scalable crossbar arrays are expected to meet the demand for high-density memory storage, fast switching speed, and ultra-low power operation. However, two-terminal crossbar arrays introduce an unintended sneak path, which inevitably requires bidirectional nonlinear selectors. In this study, the advanced threshold switching (TS) features of ZnTe chalcogenide material-based selectors provide bidirectional threshold switching behavior, nonlinearity of 104, switching speed of less than 100 ns, and switching endurance of more than 107. In addition, thermally robust ZnTe selectors (up to 400 ℃) can be obtained through the use of nitrogen-annealing treatment. This process can prevent possible phase separation phenomena observed in generic chalcogenide materials during thermal annealing which occurs even at a low temperature of 250 ℃. The possible characteristics of the electrically and thermally advanced TS nature are described by diverse structural and electrical analyses through the Poole–Frankel conduction model. [ABSTRACT FROM AUTHOR]

Oxygen-doped germanium telluride phase change materials are proposed for high temperature applications. Up to 8 at.% oxygen is readily incorporated into GeTe, causing an increased crystallisation temperature and activation energy. The rhombohedral structure of the GeTe crystal is preserved in the oxygen doped films. For higher oxygen concentrations the material is found to phase separate into GeO2 and TeO2, which inhibits the technologically useful abrupt change in properties. Increasing the oxygen content in GeTe-O reduces the difference in film thickness and mass density between the amorphous and crystalline states. For oxygen concentrations between 5 and 6 at.%, the amorphous material and the crystalline material have the same density. Above 6 at.% O doping, crystallisation exhibits an anomalous density change, where the volume of the crystalline state is larger than that of the amorphous. The high thermal stability and zero-density change characteristic of Oxygen-incorporated GeTe, is recommended for efficient and low stress phase change memory devices that may operate at elevated temperatures. [ABSTRACT FROM AUTHOR]