Affiliations: VLSI Design and Testing Center, Department of
Electrical and Computer Engineering, Democritus University of Thrace, Xanthi,
67100, Greece
Abstract: An optimization methodology that combines high-level exploration
with Register Transfer Level (RTL) design for the power-efficient estimation of
motion estimation algorithms on a system comprised by an FPGA and an external
memory is presented. Low power consumption is achieved by implementing an
optimum on-chip memory hierarchy inside the FPGA, and moving the bulk of
required memory transfers from the internal memory hierarchy instead of the
external memory. A case study of three popular multimedia kernels is performed
for a number of different FPGA devices. Comparisons among implementations with
and without this optimization, prove that great power efficiency can be
achieved while satisfying performance constraints.