Thermal and Signal Integrity Analysis of Novel 3D Crossbar Resistive Random Access Memories

Zayer, Fakhreddine; Lahbacha, Khitem; Dghais, Wael; Hamdi, Belgacem; Magistris, M. de; Maffucci, Antonio; Melnikov, A. V.

doi:10.1109/sapiw.2019.8781680

Cited by 7 publications

(4 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although several software algorithms, such as sparse mapping schemes, have been proposed to address the large number of weights [12,13], neuromorphic architectures still demand a high amount of storage. Against this background, a 3D memristor array would be the most effective hardware scheme for maximizing the area's efficiency [14][15][16]. However, the retention of written data on the device could affect the energy efficiency of the system's operation [15,16].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Long-Term Accuracy Enhancement of Binary Neural Networks Based on Optimized Three-Dimensional Memristor Array

Zhang

Dong

et al. 2022

Micromachines

View full text Add to dashboard Cite

In embedded neuromorphic Internet of Things (IoT) systems, it is critical to improve the efficiency of neural network (NN) edge devices in inferring a pretrained NN. Meanwhile, in the paradigm of edge computing, device integration, data retention characteristics and power consumption are particularly important. In this paper, the self-selected device (SSD), which is the base cell for building the densest three-dimensional (3D) architecture, is used to store non-volatile weights in binary neural networks (BNN) for embedded NN applications. Considering that the prevailing issues in written data retention on the device can affect the energy efficiency of the system’s operation, the data loss mechanism of the self-selected cell is elucidated. On this basis, we introduce an optimized method to retain oxygen ions and prevent their diffusion toward the switching layer by introducing a titanium interfacial layer. By using this optimization, the recombination probability of Vo and oxygen ions is reduced, effectively improving the retention characteristics of the device. The optimization effect is verified using a simulation after mapping the BNN weights to the 3D VRRAM array constructed by the SSD before and after optimization. The simulation results showed that the long-term recognition accuracy (greater than 105 s) of the pre-trained BNN was improved by 24% and that the energy consumption of the system during training can be reduced 25,000-fold while ensuring the same accuracy. This work provides high storage density and a non-volatile solution to meet the low power consumption and miniaturization requirements of embedded neuromorphic applications.

show abstract

Section: Introductionmentioning

confidence: 99%

“…Against this background, a 3D memristor array would be the most effective hardware scheme for maximizing the area's efficiency [14][15][16]. However, the retention of written data on the device could affect the energy efficiency of the system's operation [15,16].…”

Section: Introductionmentioning

confidence: 99%

Long-Term Accuracy Enhancement of Binary Neural Networks Based on Optimized Three-Dimensional Memristor Array

Zhang

Dong

et al. 2022

Micromachines

View full text Add to dashboard Cite

show abstract

“…In this multiphysics model (described in Section II), the electrical power dissipation is the heat source of the thermal problem, and temperature-dependent electrical parameters are considered into the electrical one. In [18] and [19], the authors have already demonstrated that issues such as voltage drop and temperature rise may appear in 3-D RRAM x-bar structures. In this article, a more detailed signal and thermal integrity analysis is provided for a large (5 × 5 × 5) structure, including the electrical effects of the thermal crosstalk.…”

Section: Introductionmentioning

confidence: 99%

“…Equivalent electrical conductivity of the bundle of MWCNT. Inset: equivalent electrical conductivity for the contact layer[18].…”

mentioning

confidence: 99%

Signal and Thermal Integrity Analysis of 3-D Stacked Resistive Random Access Memories

Zayer

Lahbacha

Melnikov

et al. 2021

IEEE Trans. Electron Devices

Self Cite

View full text Add to dashboard Cite

In this article, a 3-D electrothermal numerical model is used to perform the signal and thermal integrity analysis of 3-D stacked Resistive-switching random access memory (RRAM) arrays. Two main issues are found: voltage drop along the interconnects and thermal crosstalk between the memory cells. Possible solutions to these issues are here thoroughly investigated, based either on new biasing schemes or new materials. Especially, conventional nickel bars are replaced by interconnects made by copper (Cu) and carbon nanotubes (CNTs), whose electrical and thermal parameters are here described using physically based models. The analysis is performed on a 5 × 5 × 5 array, under a real case of a RESET switching, which is the worst case scenario from the electrothermal point of view. Simulation results show that the use of CNTs reduces the voltage drop in both word and bitline (BL) interconnects, thermal crosstalk, and the maximum working temperature; hence, it mitigates many of the crucial issues in the roadmap for the large-scale monolithic 3-D RRAM integration.

show abstract

Exploring Memristive Biosensing Dynamics: A COMSOL Multiphysics Approach

Bouzouita,

Zayer,

Tzouvadaki

et al. 2024

2024 IEEE BioSensors Conference (BioSensors)

View full text Add to dashboard Cite

Thermal and Signal Integrity Analysis of Novel 3D Crossbar Resistive Random Access Memories

Cited by 7 publications

References 11 publications

Long-Term Accuracy Enhancement of Binary Neural Networks Based on Optimized Three-Dimensional Memristor Array

Long-Term Accuracy Enhancement of Binary Neural Networks Based on Optimized Three-Dimensional Memristor Array

Signal and Thermal Integrity Analysis of 3-D Stacked Resistive Random Access Memories

Exploring Memristive Biosensing Dynamics: A COMSOL Multiphysics Approach

Contact Info

Product

Resources

About