Method for process variation analysis of an integrated circuit

A system for process variation analysis of an integrated circuit, the system comprising:

• one or more memory units configured to store a netlist describing electronic devices of an integrated circuit in terms of device parameters and process parameters of the electronic devices, wherein the process parameters include local process parameters individual to the electronic devices and global process parameters common to the electronic devices;
• a hierarchical sensitivity module configured to:
• identify critical electronic devices in the netlist as electronic devices having device parameters with greatest contributions to one or more performance parameters of a design specification;
• and
• determine sensitivity values for the global process parameters and local process parameters of the critical electronic devices, wherein the sensitivity values represent how much the one or more performance parameters change in response to variations in the global process parameters and the local process parameters of the critical electronic devices;
• a Monte Carlo (MC) sample sorting module configured to sort MC samples, wherein the MC samples are vectors of values for the process parameters, and wherein the sorting is based on the sensitivity values and the values of the MC samples;
• a MC simulation module configured to perform a MC simulation on the sorted MC samples;
• and
• a semiconductor substrate on which the integrated circuit is manufactured, wherein the integrated circuit is selectively manufacturing depending on results of the MC simulation.

A method for process variation analysis of an integrated circuit, the method comprising:

• receiving a netlist describing electronic devices of an integrated circuit;
• restructuring the netlist to include device parameters and process parameters respectively used by device and process models to model the electronic devices, wherein the process parameters include local process parameters individual to the electronic devices and global process parameters common to the electronic devices;
• identifying critical electronic devices having device parameters with greatest contributions to one or more performance parameters of a design specification;
• after identifying the critical electronic devices, determining sensitivity values individual to the global process parameters and individual to local process parameters of the critical electronic devices, wherein the sensitivity values represent how much the one or more performance parameters change in response to variations in the global process parameters and the local process parameters of the critical electronic devices;
• sorting Monte Carlo (MC) samples, wherein the MC samples are vectors of values for the process parameters, and wherein the sorting is based on the sensitivity values and the values of the MC samples;
• performing a MC simulation on the sorted MC samples, wherein the MC simulation processes MC samples proximate to one or more limits of the design specification first;
• and
• manufacturing the integrated circuit on a semiconductor substrate selectively depending upon results of the MC simulation.