Logical Reverse Inference Realized by Combined Bits with Backhopping‐Like Current‐Induced Magnetization Switching
Abstract Probabilistic computing, with its efficient and non‐deterministic computational model, holds great potential and is driving a paradigm shift in computation. However, traditional CMOS circuits face issues such as a high device count, significant energy consumption, and pseudo‐random behavior. Additionally, commonly used probabilistic computing devices like LBNM and sMTJ suffer from problems related to volatility and initialization. This paper explores the application of non‐volatile combined bit (c‐bit) devices, exhibiting back‐hopping‐like phenomena, in logical operations, arithmetic computations, and data backup and recovery. Logical reverse inferences are successfully implemented for AND, OR, XOR, and XNOR operations. Importantly, the logical reverse inference of full and half adders is achieved, as well as the decomposition of addends, verifying the feasibility of c‐bit devices in complex arithmetic operations. Furthermore, by utilizing multiple c‐bit devices working in tandem, efficient data backup and reliable recovery are achieved, offering a novel solution for data security. This research not only enriches the theoretical framework of probabilistic computing but also provides new insights and directions for the future development of computational models.
Authors
- Minggao Zuo (ORCID: https://orcid.org/0009-0001-5768-9027)
- Guanghua Yu
- Yong Hu (ORCID: https://orcid.org/0000-0001-5595-7625)
- Zeting Du
- Yan Liu (ORCID: https://orcid.org/0000-0002-9258-1003)
- Ruizhi Ren (ORCID: https://orcid.org/0000-0003-4956-3050)
- Yi Heng Cao (ORCID: https://orcid.org/0000-0001-7718-0394)
- Jiao Teng (ORCID: https://orcid.org/0000-0002-0840-876X)
- Songsong Li (ORCID: https://orcid.org/0000-0001-9784-2877)
- Jing Bao
- Chun Feng
Institutions
- Northeastern University (CN)
- University of Science and Technology Beijing (CN)
Publication Details
- Journal
- Advanced Functional Materials
- Published
- 2025-05-24
- DOI
- https://doi.org/10.1002/adfm.202501992
- Primary Topic
- Advanced Memory and Neural Computing
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- National Natural Science Foundation of China
- Natural Science Foundation of Liaoning Province
- Beijing Laboratory of Metallic Materials and Processing for Modern Transportation
- Fundamental Research Funds for the Central Universities