Processing-in-memory (PIM) is a computational architecture in which the processing unit and memory are integrated into a single unit. Different technologies and methods can be used to implement PIM, but a more optimal design for PIM can be obtained by using quantum-dot cellular automata (QCA) and Akers structure. In the chemical manufacturing process of such circuits, the manufacturing defects and therefore malfunction are possible. One of the most common defects is cell omission, which has not yet been investigated in PIM based on binary or ternary QCA. One of the aims of this study is to improve fault tolerance (FT) by transforming the circuit structure from binary to ternary QCA. The results of evaluating different structures confirm achieving this aim. Moreover, in this study, new basic PIM cells are proposed, and then, AND, OR, and Exclusive-OR (XOR) gates are designed based on these cells in ternary quantum-dot cellular automata (TQCA). The ternary structure is used to improve circuit performance, storage density, and processing capability. For better analysis and comparison, binary structures are also implemented. The results of the evaluations show that the proposed ternary structures are efficient in terms of occupied area, latency, cost, FT, and complexity, as well as better efficiency in processing and storing data because of the three-valued structure.

内存处理 (PIM) 是一种计算架构，其中处理单元和内存集成为一个单元。PIM可以采用不同的技术和方法来实现，但更优化的PIM设计可以通过使用量子点元胞自动机（QCA）和Akers结构来获得。在这种电路的化学制造过程中，制造缺陷和因此故障是可能的。最常见的缺陷之一是单元遗漏，尚未在基于二元或三元 QCA 的 PIM 中对其进行研究。本研究的目的之一是通过将电路结构从二元 QCA 转换为三元 QCA 来提高容错 (FT)。评估不同结构的结果证实实现了这一目标。此外，在本研究中，提出了新的基本 PIM 单元，然后，AND、OR、和异或（XOR）门是基于三元量子点元胞自动机（TQCA）中的这些单元设计的。三元结构用于提高电路性能、存储密度和处理能力。为了更好地分析和比较，还实现了二进制结构。评估结果表明，所提出的三元结构在占用面积、延迟、成本、FT和复杂度方面是有效的，并且由于三值结构在处理和存储数据方面的效率更高。