Intrinsic variability of nanoscale CMOS technology for logic and memory.

The continuous downscaling of CMOS technology, the main engine of development of the semiconductor Industry, is limited by factors that become important for nanoscale device size, which undermine proper device operation completely offset gains from scaling. One of the main problems is device variability: nominally identical devices are different at the microscopic level due to fabrication tolerance and the intrinsic granularity of matter. For this reason, structures, devices and materials for the next technology nodes will be chosen for their robustness to process variability, in agreement with the ITRS (International Technology Roadmap for Semiconductors). Examining the dispersion of various physical and geometrical parameters and the effect these have on device performance becomes necessary. In this thesis, I focus on the study of the dispersion of the threshold voltage due to intrinsic variability in nanoscale CMOS technology for logic and for memory. In order to describe this, it is convenient to have an analytical model that allows, with the assistance of a small number of simulations, to calculate the standard deviation of the threshold voltage due to the various contributions.