Author:
Mahadevaiah Mamathamba Kalishettyhalli,Lisker Marco,Fraschke Mirko,Marschmeyer Steffen,Perez Eduardo,Quesada Emilio Perez-Bosch,Wenger Christian,Mai Andreas
Abstract
AbstractThe two main features of the memristive devices which makes them the promising candidates for neuromorphic applications are low power consumption and CMOS compatibility. The monolithic integration of memristive devices with CMOS circuitry paves the way for in-memory computing. This chapter focuses on the factors governing the CMOS integration process. Firstly, the influence of CMOS baseline technology selection on the memristor module is briefly discussed. Secondly, the selection of metal level interconnects and their effect on the memristive device performance is explained. Further, the widely used deposition technique for the CMOS compatible memristive switching layers is presented. Finally, the implementation of the optimized process for the fabrication of the memristive module and its influence on the device performance is presented in terms of electrical characterization results.
Publisher
Springer International Publishing
Reference45 articles.
1. Yin, S., et al.: Monolithically integrated RRAM-and CMOS-based in-memory computing optimizations for efficient deep learning. IEEE Micro. 39, 54–63 (2019)
2. Staff, E.: Monolithic embedded RRAM presents challenges, opportunities kernel description. https://www.embedded.com/monolithic-embedded-rram-presents-challenges-opportunities/ (2021)
3. Zhang, W., et al.: Neuro-inspired computing chips. Nat. Electron. 3, 371–382 (2020)
4. Kim, S.G., Han, J.S., Kim, H., Kim, S.Y., Jang, H.W.: Recent advances in memristive materials for artificial synapses. Adv. Mater. Technol. 3, 1800457 (2018)
5. Mahadevaiah, M.K., Perez, E., Wenger, C.: Influence of specific forming algorithms on the device-to-device variability of memristive Al-doped HfO$$_2$$ arrays. J. Vac. Sci. & Technol. B Nanotechnol. Microelectron. Mater. Process. Meas. Phenom. 38, 013201 (2020)
Cited by
2 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献