Voltage-controlled magnetoelectric devices for neuromorphic diffusion process

Abstract Neuromorphic diffusion models have become one of the major breakthroughs in the field of generative artificial intelligence. Unlike discriminative models that have been well developed to tackle classification or regression tasks, diffusion models aim at creating content based upon contexts...

Full description

Saved in:
Bibliographic Details
Main Authors: Yang Cheng, Qingyuan Shu, Albert Lee, Haoran He, Ivy Zhu, Minzhang Chen, Renhe Chen, Zirui Wang, Hantao Zhang, Chih-Yao Wang, Shan-Yi Yang, Yu-Chen Hsin, Cheng-Yi Shih, Hsin-Han Lee, Ran Cheng, Kang L. Wang
Format: Article
Language:English
Published: Nature Portfolio 2025-05-01
Series:Nature Communications
Online Access:https://doi.org/10.1038/s41467-025-58932-x
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850243134549655552
author Yang Cheng
Qingyuan Shu
Albert Lee
Haoran He
Ivy Zhu
Minzhang Chen
Renhe Chen
Zirui Wang
Hantao Zhang
Chih-Yao Wang
Shan-Yi Yang
Yu-Chen Hsin
Cheng-Yi Shih
Hsin-Han Lee
Ran Cheng
Kang L. Wang
author_facet Yang Cheng
Qingyuan Shu
Albert Lee
Haoran He
Ivy Zhu
Minzhang Chen
Renhe Chen
Zirui Wang
Hantao Zhang
Chih-Yao Wang
Shan-Yi Yang
Yu-Chen Hsin
Cheng-Yi Shih
Hsin-Han Lee
Ran Cheng
Kang L. Wang
author_sort Yang Cheng
collection DOAJ
description Abstract Neuromorphic diffusion models have become one of the major breakthroughs in the field of generative artificial intelligence. Unlike discriminative models that have been well developed to tackle classification or regression tasks, diffusion models aim at creating content based upon contexts learned. However, the more complex algorithms of these models result in high computational costs using today’s technologies. Here, we develop a spintronic voltage-controlled magnetoelectric memory hardware for the neuromorphic diffusion process. The in-memory computing capability of our spintronic devices goes beyond current Von Neumann architecture, where memory and computing units are separated. Together with the non-volatility of magnetic memory, we can achieve high-speed and low-cost computing, which is desirable for the increasing scale of generative models in the current era. We experimentally demonstrate that the hardware-based true random diffusion process can be implemented for image generation and achieve comparable image quality to software-based training as measured by the Fréchet inception distance (FID) score, achieving ~103 better energy-per-bit-per-area over traditional hardware.
format Article
id doaj-art-e6ab33248a1241d380ce036db7337345
institution OA Journals
issn 2041-1723
language English
publishDate 2025-05-01
publisher Nature Portfolio
record_format Article
series Nature Communications
spelling doaj-art-e6ab33248a1241d380ce036db73373452025-08-20T02:00:03ZengNature PortfolioNature Communications2041-17232025-05-011611810.1038/s41467-025-58932-xVoltage-controlled magnetoelectric devices for neuromorphic diffusion processYang Cheng0Qingyuan Shu1Albert Lee2Haoran He3Ivy Zhu4Minzhang Chen5Renhe Chen6Zirui Wang7Hantao Zhang8Chih-Yao Wang9Shan-Yi Yang10Yu-Chen Hsin11Cheng-Yi Shih12Hsin-Han Lee13Ran Cheng14Kang L. Wang15Department of Electrical and Computer Engineering, University of CaliforniaDepartment of Electrical and Computer Engineering, University of CaliforniaDepartment of Electrical and Computer Engineering, University of CaliforniaDepartment of Electrical and Computer Engineering, University of CaliforniaDepartment of Physics, The Ohio State UniversityDepartment of Electrical and Computer Engineering, University of CaliforniaDepartment of Electrical and Computer Engineering, University of CaliforniaDepartment of Electrical and Computer Engineering, University of CaliforniaDepartment of Physics and Astronomy, University of CaliforniaIndustrial Technology Research InstituteIndustrial Technology Research InstituteIndustrial Technology Research InstituteIndustrial Technology Research InstituteIndustrial Technology Research InstituteDepartment of Electrical and Computer Engineering, University of CaliforniaDepartment of Electrical and Computer Engineering, University of CaliforniaAbstract Neuromorphic diffusion models have become one of the major breakthroughs in the field of generative artificial intelligence. Unlike discriminative models that have been well developed to tackle classification or regression tasks, diffusion models aim at creating content based upon contexts learned. However, the more complex algorithms of these models result in high computational costs using today’s technologies. Here, we develop a spintronic voltage-controlled magnetoelectric memory hardware for the neuromorphic diffusion process. The in-memory computing capability of our spintronic devices goes beyond current Von Neumann architecture, where memory and computing units are separated. Together with the non-volatility of magnetic memory, we can achieve high-speed and low-cost computing, which is desirable for the increasing scale of generative models in the current era. We experimentally demonstrate that the hardware-based true random diffusion process can be implemented for image generation and achieve comparable image quality to software-based training as measured by the Fréchet inception distance (FID) score, achieving ~103 better energy-per-bit-per-area over traditional hardware.https://doi.org/10.1038/s41467-025-58932-x
spellingShingle Yang Cheng
Qingyuan Shu
Albert Lee
Haoran He
Ivy Zhu
Minzhang Chen
Renhe Chen
Zirui Wang
Hantao Zhang
Chih-Yao Wang
Shan-Yi Yang
Yu-Chen Hsin
Cheng-Yi Shih
Hsin-Han Lee
Ran Cheng
Kang L. Wang
Voltage-controlled magnetoelectric devices for neuromorphic diffusion process
Nature Communications
title Voltage-controlled magnetoelectric devices for neuromorphic diffusion process
title_full Voltage-controlled magnetoelectric devices for neuromorphic diffusion process
title_fullStr Voltage-controlled magnetoelectric devices for neuromorphic diffusion process
title_full_unstemmed Voltage-controlled magnetoelectric devices for neuromorphic diffusion process
title_short Voltage-controlled magnetoelectric devices for neuromorphic diffusion process
title_sort voltage controlled magnetoelectric devices for neuromorphic diffusion process
url https://doi.org/10.1038/s41467-025-58932-x
work_keys_str_mv AT yangcheng voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess
AT qingyuanshu voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess
AT albertlee voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess
AT haoranhe voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess
AT ivyzhu voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess
AT minzhangchen voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess
AT renhechen voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess
AT ziruiwang voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess
AT hantaozhang voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess
AT chihyaowang voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess
AT shanyiyang voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess
AT yuchenhsin voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess
AT chengyishih voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess
AT hsinhanlee voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess
AT rancheng voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess
AT kanglwang voltagecontrolledmagnetoelectricdevicesforneuromorphicdiffusionprocess