Simulation of Statistical Variability in Nano MOSFETs

Using 3D statistical numerical simulations we study and compare the impact of various sources of statistical variability in nano-CMOS transistors including random discrete dopants (RDD), line edge roughness (LER), polysilicon granularity (PSG) and interface roughness (IR). We show that the random dopant induced parameter fluctuations in conventional MOSFETs become a showstopper, impacting already adversely on their integration in SRAMs, and may force early transition to UTB SOI and double gate device architectures.