XBarNet: Computationally Efficient Memristor Crossbar Model Using Convolutional Autoencoder

The design and verification of memristor crossbar circuits and systems demand computationally efficient models. The conventional device-level memristor model with a circuit simulator such as simulation program with integrated circuit emphasis (SPICE) to solve a memristor crossbar is time exhaustive....

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IEEE transactions on computer-aided design of integrated circuits and systems Ročník 41; číslo 12; s. 5489 - 5500
Hlavní autoři: Zhang, Yuhang, He, Guanghui, Wang, Guoxing, Li, Yongfu
Médium: Journal Article
Jazyk:angličtina
Vydáno: New York IEEE 01.12.2022
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Témata:
Circuit design
Circuit simulation
Computational efficiency
Convolutional autoencoder
Feature extraction
Integrated circuit interconnections
Integrated circuit modeling
Integrated circuits
interconnect
Iterative methods
memristor crossbar
Memristors
modeling
Modelling
Neural networks
nonlinearity
Pixels
Run time (computers)
Simulation
ISSN:0278-0070, 1937-4151
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:The design and verification of memristor crossbar circuits and systems demand computationally efficient models. The conventional device-level memristor model with a circuit simulator such as simulation program with integrated circuit emphasis (SPICE) to solve a memristor crossbar is time exhaustive. Hence, we propose a neural network-based memristor crossbar modeling method, XBarNet. By transforming memristor crossbar modeling to pixel-to-pixel regression, XBarNet avoids the iterative procedure in the conventional SPICE method, accelerating the runtime significantly. Meanwhile, XBarNet models the interconnect resistance and nonlinear <inline-formula> <tex-math notation="LaTeX">I-V </tex-math></inline-formula> effect of memristor crossbars, which minimizes the simulation errors. We first propose a feature extraction method to bridge a memristor crossbar circuit and a neural network. Then, the network based on the convolutional autoencoder architecture is developed and the filter pruning technique is applied onto XBarNet to reduce the runtime computational cost. The experimental result shows our proposed XBarNet achieves over <inline-formula> <tex-math notation="LaTeX">78\times </tex-math></inline-formula> runtime speed up and <inline-formula> <tex-math notation="LaTeX">1.7\times </tex-math></inline-formula> memory reduction with only 0.28% relative error comparing to the SPICE simulator.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0278-0070
1937-4151
DOI:10.1109/TCAD.2022.3163895