Static Timing IR Drop Analysis Using Fuzzy Bins

The patent segments the simulation process by dividing the timing signal into multiple bins, grouping devices based on their activation time windows. This segmentation allows the simulation to consider only devices that are simultaneously active within each bin, rather than assuming all devices are always at maximum current. The segmentation principle resolves the contradiction by enabling accurate IR testing through time-based grouping while maintaining computational efficiency.

The patent applies dynamics by transitioning from a static assumption (all devices simultaneously at maximum current) to a dynamic model where device activation is time-dependent. By using bin-based time windows and tracking device activation states dynamically, the simulation accurately reflects real-world conditions where devices activate at different times. This dynamic approach improves measurement precision while the binning strategy keeps the computational complexity manageable.

The patent applies preliminary action by pre-defining bin structures and device activation time windows before performing the IR analysis. Devices are pre-grouped into bins based on their expected activation times, and this grouping is used throughout the simulation. This preliminary organization enables efficient calculation by avoiding the need to check all device combinations, while still achieving accurate results by considering only simultaneously active devices within each bin.

The patent changes the parameter of time by introducing bin-based time windows with associated duration parameters. Instead of using a single worst-case assumption, the simulation varies the time parameter across multiple bins, allowing devices to be evaluated at different activation times. This parameter change enables accurate current demand calculation by reflecting temporal variations in device activation while maintaining computational efficiency through the binning structure.