高精度阻变存算一体矩阵芯片:基于F_q‑Fₑ联合模型的粒子确定态溯源分析
Abstract Resistive Random-Access Memory (RRAM) computing-in-memory architectures effectively break the inherent von Neumann bottleneck of traditional chips. Recent advanced precision analog matrix solving chips significantly suppress internal conductance drift of resistive units, achieving substantial improvements in throughput and energy efficiency for large-scale MIMO signal detection and matrix inversion tasks. Although conventional device physics can qualitatively interpret experimental phenomena based on oxygen vacancy conductive filaments and carrier transport mechanisms, it lacks a unified phenomenological framework to describe the collective behavior of multi-carrier systems. Based on the fully verified F_q-F_e joint model (short-range repulsion and medium-range attraction force model combined with slice-trajectory-correlation theory) permanently archived on Zenodo, this paper conducts a posteriori traceability analysis of RRAM computing-in-memory devices from the perspective of particle determined state. This study constructs a complete theoretical spectrum for two mainstream high-performance chip architectures: the τ-law logical folding architecture reduces particle disturbance by shortening the signal transmission path, while the RRAM computing-in-memory architecture eliminates disturbance by removing cross-module data migration paths. Strictly distinguishing experimental facts, phenomenological deductions, and underlying physical conjectures, this work clarifies the application boundaries of the model and proposes falsifiable research directions. It further expands the application scope of the particle determined state system from microscopic double-slit experiments and three-dimensional stacked chips to emerging resistive computing-in-memory devices. 摘要 阻变存储器(RRAM)存算一体架构有效突破了传统芯片固有的冯·诺依曼瓶颈。新一代高精度模拟矩阵求解芯片通过深度抑制阻变单元电导随机漂移,在大规模MIMO信号检测、矩阵求逆等专属运算场景中,实现了芯片吞吐量与能效的跨越式提升。现有器件物理理论可基于氧空位导电细丝、载流子输运机理定性解释相关实验现象,但缺乏一套能够描述多载流子集体行为的统一唯象分析框架。本文依托Zenodo平台已完整存证的F_q-Fₑ联合模型(短程排斥-中程吸引作用力模型+切片-轨迹-关联理论),从粒子确定态视角,对阻变存算一体芯片开展系统性后验溯源研究。构建两类主流高性能芯片架构的统一理论谱系:韬定律逻辑折叠架构以“缩短信号传输路径”压制粒子扰动,RRAM存算一体架构以“消除跨模块数据迁移路径”根除扰动源头。研究严格区分实验事实、唯象推演与底层物理猜想,明确模型适用边界,提出具备可证伪性的后续研究方向,将粒子确定态理论体系的应用场景,从微观双缝实验、三维堆叠芯片,进一步拓展至新型阻变存算器件领域。
Authors
- ZhenFeng BIAN (ORCID: https://orcid.org/0009-0003-7300-6374)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-18
- DOI
- https://doi.org/10.5281/zenodo.24329945
- Primary Topic
- Advanced Memory and Neural Computing
- Type
- preprint