Multiobjective Floorplanning for Partially-Reconfigurable FPGA Systems

The aim of this work is to present a novel floorplanner tailored for partially reconfigurable FPGAs, taking into account an accurate description of heterogeneous resources and partial reconfiguration constraints. The proposed approach is based on a suitable Mixed-Integer Linear Programming (MILP) formulation that makes the problem tractable using state-of-the-art solvers. A global optimum can be found for small instances in the short run. For large instances, with a time limited search, a 20% average improvement can be achieved over state-of-the-art floorplanners. The described methodology allows the designer to customize the objective function to be minimized, so that different weights can be assigned to a linear combination of cost functions such as total wirelength, area occupancy and desired aspect ratio.