"Nowadays the possibility to store and exchange large amount of data is driving the technology research. Every single year we produce more than 50 billion Gb of memory and this rate is increasing very quickly.In recent years a new class of memory is gaining a lot of interest in the electronics world. It is called Resistive Random Access Memory (abbreviated as RRAM) and it includes a large variety of memory concepts. The common denominator is that all these devices store the information as a resistance value of a device. The resistance can be changed reversibly by applying a voltage across the device multiple times. RRAM is very fast (write and erase can be achieved with less than 1ns) and work with very low voltages (in some cases as low as 1V). This makes these memories ideal candidates for applications like IoT, or DRAM replacement.RRAM show however a peculiar issue. The memory working principle relies on the formation and destruction of tiny conductive filaments across an insulating material. This filament is made by a very limited number of atoms that could be as low as a few tens. If one electron is captured by an atom, the entire filament will be sensitive to the electron presence, and the filament resistance will change. In this way we can “see” the effect of single electrons.The stage will be divided into two main parts. First, the characterization of this resistance fluctuation on memory prototypes and then the modeling of this phenomenon from a statistical point of view.The aim of the stage is to quantify the impact of this fluctuation on the memory performance and offer a better physical picture of it."