Investigation of statistical retention of filamentary analog RRAM for neuromophic computing

Zhao, Meiran; Wu, Huaqiang; Gao, Bin; Zhang, Qingtian; Wu, Wei; Wang, Shan; Xi, Yue; Wu, Dong; Deng, Ning; Yu, Shimeng; Chen, Hongyu; Qian, He

doi:10.1109/iedm.2017.8268522

Cited by 76 publications

(47 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[124][125][126][127] As mentioned earlier, the insufficient dynamic range and poor weight update symmetry are the critical obstacles to improving the performance of nonvolatile memory-based neuromorphic computing. [128][129][130] Meanwhile, many performances offered by complex PCM network architectures were matchable or superior to that of the software platform, particularly in power consumption, classification accuracy, and training speed. The end-to-end training approaches of multilayer perceptrons need parallel update of all-synaptic weights during one training epoch by sending overlapping pulses from columns or rows.…”

Section: Electrical Systems and Applicationsmentioning

confidence: 99%

Exploring Phase‐Change Memory: From Material Systems to Device Physics

Ren

Sun

Chen³

et al. 2021

Physica Rapid Research Ltrs

View full text Add to dashboard Cite

To deal with the growing demand for data storage and processing, phase‐change memory (PCM) provides one of the most promising candidates for next‐generation nonvolatile data storage and neuromorphic computing applications. A lot of effort has been made toward optimizing the materials and device design; thus, excellent device performances have been achieved including high density, fast switching speed, great endurance, and retention. In addition, the widely tunable optical characteristics of PCMs are irresistibly attractive for optoelectronic or all‐optical applications with unprecedented bandwidth, low energy consumption, and multilevel data storage. Herein, the materials system and switching mechanisms on experimental and modeling methods for PCM designs and applications are discussed. Electric‐domain and optical‐domain PCM‐based artificial synapses/neurons and their applications in neuromorphic computing are also reviewed. Finally, the future prospects and challenges of PCM‐based applications on materials, devices, algorithms, and system levels are highlighted.

show abstract

Section: Electrical Systems and Applicationsmentioning

confidence: 99%

Exploring Phase‐Change Memory: From Material Systems to Device Physics

Ren

Sun

Chen³

et al. 2021

Physica Rapid Research Ltrs

View full text Add to dashboard Cite

show abstract

“…For practical applications, the retention can be further improved by using our optimized HfOx-based analog RRAM. The Al-doping HfOx layer contributes to significant retention improvement which has been systematically studied in [33]. The multi-bit data can be reliably stored for more than 8 years at 85 degree.…”

Section: Inter-hdmentioning

confidence: 99%

A Unified Memory and Hardware Security Module Based on the Adjustable Switching Window of Resistive Memory

Lin

Gao

Pang

et al. 2020

IEEE J. Electron Devices Soc.

Self Cite

View full text Add to dashboard Cite

Physically unclonable function (PUF) and true random number generator (TRNG) are critical primitives to provide lightweight hardware protection. As area is tightly restricted in IoT applications, merging PUF and TRNG is a novel trend to achieve higher area efficiency. In this work, a novel tri-functional module is proposed and experimental demonstrated using RRAM for the first time. A two-phase forming process is designed to generate and store PUF ID utilizing the impact of different forming conditions on the switching window. As PUF ID is stored by the switching window, instead of a fixed resistance, PUF cells can be written to low resistance state and high resistance state as usual, and can be used as memory and TRNG. The inter-Hamming distance of the generated PUF IDs approaches 0.5, and with the majority voting readout strategy, low BER can be achieved in a wide range of temperatures from-40 o C to 125 o C. TRNG produces the true random numbers based on the parity of the number of pulses consumed in a write-zero process. The generated random numbers are uniform and uncorrelated after XORed, and the statistical randomness is verified by NIST SP800-22 tests.

show abstract

“…It can be verified that the slope of the curve is the reciprocal of the standard deviation (σ ) of the normal distribution. In our previous work, σ acts as a function of the square root of baking time ( √ t) within 600s at 125 • C [19]. Then with studying the retention degradation behavior in a longer time span at higher temperature, the relationship between σ and baking time are presented more clearly in Eq.…”

Section: A Retentionmentioning

confidence: 99%

“…In this formula, η 0 is the vibration frequency of oxygen ions; E h is the hopping barrier of oxygen ions, which is the intrinsic parameter depending on the material and structure of device. It could be extracted by referring to the method based on the Arrhenius equation in [19]. t is a unit retention time; k is the Boltzmann constant.…”

Section: Physical Mechanismsmentioning

confidence: 99%

Endurance and Retention Degradation of Intermediate Levels in Filamentary Analog RRAM

Zhao

Gao

et al. 2019

IEEE J. Electron Devices Soc.

Self Cite

View full text Add to dashboard Cite

The understanding of the retention and endurance degradation behavior of different levels of filamentary analog RRAM is critical for the development of the neuromorphic computing. This paper investigates the conductance distribution of different levels during retention and endurance tests. The low conductance states and high conductance states change from the normal distribution at the beginning to asymmetric skewed distribution with baking time increasing. But intermediate states remain the normal distribution. A model is proposed to predict the conductance evolution of different levels. It is also found that endurance lifetime depends on the ratio and position of endurance window. The longer endurance lifetime is attributed to switching in a smaller high-C window. Physical mechanism of endurance degradation is discussed.

show abstract

Investigation of statistical retention of filamentary analog RRAM for neuromophic computing

Cited by 76 publications

References 7 publications

Exploring Phase‐Change Memory: From Material Systems to Device Physics

Exploring Phase‐Change Memory: From Material Systems to Device Physics

A Unified Memory and Hardware Security Module Based on the Adjustable Switching Window of Resistive Memory

Endurance and Retention Degradation of Intermediate Levels in Filamentary Analog RRAM

Contact Info

Product

Resources

About