Chip‐Level Voltage Controlled Magnetic p‐Bits Array with Variation Compensation for Stochastic Neural Network Computation
Advanced Science · 10.11.2026
DOI: 10.1002/advs.78604 ↗Kurz gefasst
Fachartikel in der Zeitschrift Advanced Science, online erschienen am 10.11.2026. Verfasst von Jiachen Bao und 15 weiteren Autor:innen. Beteiligte Einrichtung: School of Materials Science and Engineering, University of Science and Technology Beijing, Beijing, China. Die Arbeit nennt 59 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Natural Science Foundation of China, National Key Research and Development Program of China, Fundamental Research Funds for the Central Universities of the Ministry of Education of China.
Abstract (Originalsprache)
The increasing computational and energy‐efficiency demands of artificial intelligence are driving the development of new computing paradigms based on emerging memory devices. Probabilistic bits (p‐bits) based on magnetic tunnel junctions (MTJs), benefiting from energy efficiency, CMOS compatibility, and fast sampling, show broad potential in stochastic neural networks. However, unavoidable device‐to‐device variations in MTJ switching probabilities hinder the realization of scalable p‐bit chips. Here, we integrate a high‐density array of voltage‐controlled magnetic anisotropy MTJs (VCMA‐MTJs) and employ a tail truncation strategy to compensate for their variations in probabilistic switching ranges, enabling chip‐level p‐bit functionality. Our VCMA‐MTJ chip (VM‐chip) successfully implements restricted Boltzmann machines (RBM) and deep belief networks (DBN) for image recognition, anomaly detection, image completion, and image generation, demonstrating the practical utility of this chip in stochastic neural networks. Furthermore, we propose a dual‐control strategy that tunes both the amplitude and duration of the voltage pulse, significantly enhancing network performance and performing comparably to conventional software‐generated pseudo‐random numbers. Our work enhances the tolerance of probabilistic computing to variations in MTJ switching probabilities, offering flexibility in device selection and array integration. This approach may provide a promising pathway toward scalable p‐bit hardware and cost‐effective manufacturing in future probabilistic computing systems.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- School of Materials Science and Engineering, University of Science and Technology Beijing, Beijing, China.
- College of Sciences, Northeastern University, Shenyang, China.
- Institute of Integrated Circuits, Shanghai University, Shanghai, China.
- Key Laboratory of Advanced Materials and Devices for Post-Moore Chips, Ministry of Education, University of Science and Technology Beijing, Beijing, China.
ORCID-Profile: Jiachen Bao, Zheng Zhu, Jiayong Wang, Minggao Zuo, Chuyue Liu, Wangyang Hu, Yi Cao
Förderung
- Natural Science Foundation of China
- National Key Research and Development Program of China
- Fundamental Research Funds for the Central Universities of the Ministry of Education of China
- National Natural Science Foundation of China
Zitieren
APA
Bao, J., Li, S., Ren, R., Zhu, Z., Wu, D., Wang, J., Zuo, M., Li, M., Cheng, D., Lv, J., Zou, L., Liu, C., Hu, W., Cao, Y., Li, Y., & Yu, G. (2026). Chip‐Level Voltage Controlled Magnetic p‐Bits Array with Variation Compensation for Stochastic Neural Network Computation. Advanced Science. https://doi.org/10.1002/advs.78604
Vancouver
Bao J, Li S, Ren R, Zhu Z, Wu D, Wang J, et al. Chip‐Level Voltage Controlled Magnetic p‐Bits Array with Variation Compensation for Stochastic Neural Network Computation. Advanced Science. 2026. doi:10.1002/advs.78604
BibTeX
@article{Bao2026,
author = {Bao, Jiachen and Li, Songsong and Ren, Ruizhi and Zhu, Zheng and Wu, Di and Wang, Jiayong and Zuo, Minggao and Li, Mingke and Cheng, Duanzihua and Lv, Jianyu and Zou, Lintong and Liu, Chuyue and Hu, Wangyang and Cao, Yi and Li, Yuesheng and Yu, Guanghua},
title = {Chip‐Level Voltage Controlled Magnetic p‐Bits Array with Variation Compensation for Stochastic Neural Network Computation},
journal = {Advanced Science},
year = {2026},
doi = {10.1002/advs.78604}
}
EUVIMED