Optimized programming algorithms for multilevel RRAM in hardware neural networks
- A key requirement for RRAM in neural network accelerators with a large number of synaptic parameters is the multilevel programming. This is hindered by resistance imprecision due to cycle-to-cycle and device-to-device variations. Here, we compare two multilevel programming algorithms to minimize resistance variations in a 4-kbit array of HfO 2 RRAM. We show that gate-based algorithms have the highest reliability. The optimized scheme is used to implement a neural network with 9-level weights, achieving 91.5% (vs. software 93.27%) in MNIST recognition.
Author: | Valerio Milo, Francesco Anzalone, Cristian Zambelli, Eduardo Perez, Mamathamba Kalishettyhalli Mahadevaiah, Óscar G. Ossorio, Piero Olivo, Christian WengerORCiD, Daniele Ielmini |
---|---|
DOI: | https://doi.org/10.1109/IRPS46558.2021.9405119 |
ISBN: | 978-1-7281-6894-4 |
ISSN: | 1938-1891 |
Title of the source (English): | IEEE International Reliability Physics Symposium (IRPS), 2021 |
Document Type: | Conference publication peer-reviewed |
Language: | English |
Year of publication: | 2021 |
Tag: | Multilevel switching; RRAM; memristive switching; neural network |
Faculty/Chair: | Fakultät 1 MINT - Mathematik, Informatik, Physik, Elektro- und Informationstechnik / FG Halbleitermaterialien |