Resistive memory-based neural differential equation solver for score-based diffusion model

While AI-generated content (AIGC) strives to replicate human imagination, current models like score-based diffusion remain slow and energy-intensive. This inefficiency stems from conventional digital computers, where physically separated storage and processing units cause data-transfer bottlenecks, and discrete operations disrupt naturally continuous generation dynamics. Here we show a brain-inspired, analog in-memory computing system that overcomes these limitations. By employing resistive memory, our system integrates storage and computation to act as a time-continuous neural differential equation solver. We experimentally validate our solution with 180 nm resistive memory in-memory computing macros. While maintaining generative quality equivalent to the software baseline, our system accelerated both unconditional and conditional generation tasks, by factors of 69.0 and 116.5, respectively, compared to advanced digital hardware. Furthermore, it reduced energy consumption by 31.5% and 52.0%, respectively. Our approach expands the horizon for hardware solutions in edge computing for generative AI applications.

Authors

Institutions

Publication Details

Journal
Nature Communications
Published
2026-05-12
DOI
https://doi.org/10.1038/s41467-026-72900-z
Citations
1
Primary Topic
Neural Networks and Applications
Type
preprint
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
preprint

Resistive memory-based neural differential equation solver for score-based diffusion model

Dashan Shang, Songqi Wang, Jia Chen, Hegan Chen et al.
1 citations
Nature Communications
Neural Networks and Applications
preprint

Resistive memory-based neural differential equation solver for score-based diffusion model

Dashan Shang, Songqi Wang, Jia Chen, Hegan Chen, Bin‐Bin Cui, Kwang‐Ting Cheng, Shaocong Wang, Zhongrui Wang, Ming Liu, Yi Li, Han Wang, Qi Liu, Ning Lin, Jichang Yang, Xi Chen, Xiaojuan Qi, Yifei Yu, Xiaoxin Xu, Xumeng Zhang, Bo Wang, Meng Xu, Xinyuan Zhang
preprint en
1 citations

Abstract

While AI-generated content (AIGC) strives to replicate human imagination, current models like score-based diffusion remain slow and energy-intensive. This inefficiency stems from conventional digital computers, where physically separated storage and processing units cause data-transfer bottlenecks, and discrete operations disrupt naturally continuous generation dynamics. Here we show a brain-inspired, analog in-memory computing system that overcomes these limitations. By employing resistive memory, our system integrates storage and computation to act as a time-continuous neural differential equation solver. We experimentally validate our solution with 180 nm resistive memory in-memory computing macros. While maintaining generative quality equivalent to the software baseline, our system accelerated both unconditional and conditional generation tasks, by factors of 69.0 and 116.5, respectively, compared to advanced digital hardware. Furthermore, it reduced energy consumption by 31.5% and 52.0%, respectively. Our approach expands the horizon for hardware solutions in edge computing for generative AI applications.

Nature Communications
Hong Kong University of Science and Technology (HK), Huazhong Agricultural University (CN), Fudan University (CN), Southern University of Science and Technology (CN), Institute of Microelectronics (SG), Integrated Sensors (United States) (US), Institute of Microelectronics (CN), Peking University Shenzhen Hospital (CN), University of Chinese Academy of Sciences (CN), Huazhong University of Science and Technology (CN), University of Hong Kong (HK)
National Natural Science Foundation of China, Innovation and Technology Fund
Openalex Percentile: Top 100%
Neural Networks and Applications
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.