Performance Analysis of Graphene Nanoribbon as VLSI Interconnects

Abstract

The present work is an effort to find a solution for the problem faced by copper material in deep submicron technology node. With the advancement in technology, the resistivity of copper interconnects is increasing steeply due to surface scattering and grain boundary effect. Carbon nanomaterials like (Carbon nanotube) CNT and (Graphene Nanoribbon) GNR are considered as the attractive candidates for future VLSI interconnects due to large mean free paths and its capability to conduct high current density and high thermal conductivity. In this thesis report, Performance of MLGNR interconnects has been analyzed at 22 nm technology node. Similar analysis is carried out for copper interconnect and results are compared with MLGNR for local, semi-global and global interconnects. SPICE simulation results reveal that MLGNR interconnects perform better than copper for local, semi-global and global interconnects. The influence of number of layers on MLGNR interconnect performance in terms of propagation delay, power dissipation and power delay product is analyzed and it has been seen that the performance improves for more number of layers. The impact of interlayer resistance on performance of MLGNR interconnect is analyzed. The dependence of Fermi energy on impedance parameters is also critically analyzed and its effect on performance of MLGNR is examined. Finally, the effect of increase in contact resistance on propagation delay is analyzed for different lengths in MLGNR interconnect for global lengths.