The Flash memory was conceived as an improvement of the EPROM (Erasable Programmable Read Only Memory) concept invented in 1980s from an initial idea of Frohman-Bentchkowsky [1]. The EPROM memory electrically programmed and erased by ultraviolet (UV) - irradiation became the most important non-volatile memory (NVM) application in the 1980s. The Flash, which owes its name to the fact that the whole memory array can be erased quickly in the same time, introduced the advantages of the electrical erase and the possibility to reprogram the read only memory in situ, with no need of removing it from the system [2,3]. Over the years Flash memory has widely been accepted as the NVM of choice for many applications and today the large majority of NVMs is based on Flash technology. The Flash market has grown in a very fast way due to the large diffusion of portable and low power consumption multi-media applications, which requires an extensive use of NVMs. The continuous scaling of nonvolatile memories has pushed the Flash technology toward its limits [4]. Now, several constraints, mainly due to electrical and reliability reasons, are threatening the future scaling of the Flash technology and new concepts of non-volatile memories have been proposed. NVMs based on natural traps in dielectrics (such as SONOS) or on floating nanocrystals (NCs), artificially embedded in dielectrics, offer an interesting scaling alternative to the Flash with the conventional floating gate, because of several potential advantages associated with the discrete nature of the storage [5-7]. These memories are an evolution of the Flash concept where the …