arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1607.05065cs.ET
收藏

用量子点元胞自动机处理内存实现

Processing In-memory realization using Quantum Dot Cellular Automata

P. P. Chougule, B. Sen, T. D. Dongale

摘要:

本文讨论了利用量子点元胞自动机(QCA)和Akers阵列实现内存处理(PIM)计算体系结构。PIM计算架构由于其在单个单元中存储和计算数据的有效框架而变得流行。在这里,我们说明了两个输入NAND和“或非”门的帮助下,基于QCA的Akers阵列作为一个案例研究。QCA触发器被用作PIM结构设计的原始单元。结果表明,两种栅极都具有最小的功耗。极化结果表明,所提出的架构的信号是在良好的控制。原始单元的基底面积等于0.04 μ m^2,比传统CMOS原始单元小。QCA和Akers阵列的结合提供了许多额外的好处,在传统的架构,如降低功耗和特征尺寸,此外,它还提高了计算速度。

英文摘要:

The present manuscript deals with the realization of Processing In-memory (PIM) computing architecture using Quantum Dot Cellular Automata (QCA) and Akers array. The PIM computing architecture becomes popular due to its effective framework for storage and computation of data in a single unit. Here, we illustrate two input NAND and NOR gate with the help of QCA based Akers Array as a case study. The QCA flip-flop is used as a primitive cell to design PIM architecture. The results suggested that both the gate have minimum power dissipation. The polarization results of proposed architecture suggested that the signals are in good control. The foot print of the primitive cell equals to 0.04 micron^2, which is smaller than conventional CMOS primitive cell. The combination of QCA and Akers array provides many additional benefits over the conventional architecture like reduction in the power consumption and feature size, furthermore, it also improves the computational speed.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑