|MUÑOZ MARTÍN IRENE||Cycle: XXXIII |
Tutor: SOTTOCORNOLA SPINELLI ALESSANDRO Major Research topic
:Modeling and integration of resistive switching devices
Advisor: IELMINI DANIELEAbstract:
Recently, resistive switching memory (RRAM) has emerged as a promising alternative to flash memories for high-density data storage, improved performance and scalability. They have also a leading position as potential synaptic elements in large networks implementing bio-realistic plasticity rules, becoming a key element in the design of neuromorphic networks that try to mimic the event-driven spikes of human brain during cognitive processes.
The development of integrated circuits based on the neuromorphic protocol STDP (Spike-Time Dependent Plasticity) allows exploiting the full potential of RRAM, introducing them in a low power and low area-consuming network.
In order to introduce these devices at integrate level using Cadence Virtuoso, it is necessary to describe their physical properties creating new models programmed in Verilog-A and Matlab, also improving the preceding analytical models developed. HfO2 and SiOx based RRAM are the ones studied, being the latter one a promising memory element due to its improved characteristics with lower resistance variability and reduce noise.