Hardware-software partitioning is a crucial step in the design of embedded
reconfigurable systems. Typically the objective is to optimise the speed pe
rformance of the application while minimising cost. Recently, FPGAs are bei
ng employed in embedded systems to increase the computational power of the
system by customising the reconfigurable platform to the hardware partition
's requirements. A new method for hardware-software partitioning is present
ed, which considers partitioning applications within resource-limited embed
ded systems that utilise the runtime reconfiguration capabilities of FPGAs.
The partitioning approach is demonstrated using example speech and image p
rocessing applications.