Parametric macromodels for efficient design of carbon nanotube interconnects

The continuous increase of the operating frequency and density of integrated circuits leads to consider single-wall carbon nanotubes (SWCNTs) and multiwall carbon nanotubes (MWCNTs) as the most promising candidates for future interconnect technology because of their high current-carrying capacity and conductivity in the nanoscale, and immunity to electromigration. Several modeling methods for SWCNT and MWCNT interconnects have been based on the multiconductor transmission line (MTL) theory. These methods are limited to nanostructures with predefined values of electrical and geometrical parameters. Since process technology continues to scale downward and physical interconnect dimensions become smaller, the impact of design parameters (e.g., layout features) on the system behavior has to be carefully investigated for a successful design by performing design space exploration, optimization, and variability analysis. These design activities require multiple system simulations for different values of design parameters, and using MTL-based solvers is not an efficient choice. Parametric macromodels can be used to accurately and efficiently model these parameter effects, avoiding the brute-force use of MTL-based solvers.