FPGA Packing via Rent Characteristic Optimization

The patent applies segmentation by dividing the netlist into multiple partitions through recursive bipartitioning before packing into clusters. This pre-segmentation optimizes the Rent characteristic by ensuring that partitions have balanced connectivity properties, which reduces the number of external nets that need to be routed between clusters, thereby reducing overall routing requirements while maintaining implementation feasibility.

The patent performs preliminary action by conducting recursive bipartitioning of the netlist before the actual packing step. This preliminary partitioning optimizes the Rent characteristic in advance, creating a structured division of logic elements that facilitates more efficient routing in subsequent steps, thus reducing routing requirements before the packing process even begins.

The patent applies dynamics by using recursive bipartitioning that dynamically adjusts the partitioning process based on the Rent characteristic optimization goals. Rather than applying a fixed aggressive k-way partitioning, the method adaptively refines partitions through multiple levels of bipartitioning, allowing the process to maintain solution quality while achieving computational efficiency through the recursive nature of the algorithm.

The patent changes parameters by optimizing for Rent characteristic rather than simply minimizing cuts. This parameter change in the optimization objective leads to partitions that are less disruptive to initial groupings while maintaining partitioning speed, as the Rent-based optimization naturally preserves meaningful connections between logic elements that should remain together.