Binarized Neural Network Comprising Quasi‐Nonvolatile Memory Devices for Neuromorphic Computing

This study presents a binarized neural network (BNN) comprising quasi‐nonvolatile memory (QNVM) devices that operate in a positive feedback loop mechanism and exhibit an extremely low subthreshold swing (≤ 5 mV dec−1) and a high on/off ratio (≥ 107). A pair of QNVM devices are used for a single syna...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Advanced electronic materials Ročník 10; číslo 9
Hlavní autoři: Shin, Yunwoo, Jeon, Juhee, Cho, Kyoungah, Kim, Sangsig
Médium: Journal Article
Jazyk:angličtina
Vydáno: Seoul John Wiley & Sons, Inc 01.09.2024
Wiley-VCH
Témata:
Accuracy
Arrays
binarized neural network
Energy consumption
Feedback loops
image recognition
Memory devices
multiply‐accumulate
Neural networks
Neuromorphic computing
Positive feedback
positive feedback loop
quasi‐nonvolatile memory
Silicon
ISSN:2199-160X, 2199-160X
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:This study presents a binarized neural network (BNN) comprising quasi‐nonvolatile memory (QNVM) devices that operate in a positive feedback loop mechanism and exhibit an extremely low subthreshold swing (≤ 5 mV dec−1) and a high on/off ratio (≥ 107). A pair of QNVM devices are used for a single synaptic cell in a cell array, in which its memory state represents the synaptic weight, and the voltages applied to the pair act as input in a complementary fashion. The array of synaptic cells performs matrix multiply‐accumulate (MAC) operations between the weight matrix and input vector using XNOR and current summation. All the results of the MAC operations and vector‐matrix multiplications are equivalent. Moreover, the BNN features a high accuracy of 93.32% in the MNIST image recognition simulation owing to high device uniformity (1.35%), which demonstrates the feasibility of compact and high‐performance neuromorphic computing. In this study, quasi‐nonvolatile memory devices with the p+‐n‐p‐n+ structure are proposed as synaptic devices. The 2×2 synaptic cell array demonstrates the matrix multiply‐accumulate operations using XNOR and current summations. Furthermore, binarized neural network achieves high accuracy (93.32%) in MNIST image recognition simulation. Consequently, the devices provide synaptic cell array for high‐performance compact neuromorphic computing architecture.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:2199-160X
2199-160X
DOI:10.1002/aelm.202400061