Statistical variations in 32nm thin-body SOI devices and SRAM cells

Based on 3D statistical device simulation, the impacts of key statistical variability (SV) sources (in both individual and combined forms) on device characteristics are studied in detail for a 32 nm thin-body SOI technology. The corresponding impacts on SRAM cell stability are presented as well. The simulation results indicate that thin body architectures are not only resistant to random discreet dopant induced variation, but also less sensitive to length edge roughness induced variation.