Peak power estimation using genetic spot optimization for large VLSI circuits

Estimating peak power involves optimization of the circuit's switching function. We propose genetic spot expansion and optimization in this paper to estimate tight peak power bounds for large sequential circuits. The optimization spot shifts and expands dynamically based on the maximum power potential (MPP) of the nodes under optimization. Four genetic spot optimization heuristics are studied for sequential circuits. Experimental results showed an average of 70.7% tighter peak power bounds for large sequential benchmark circuits was achieved in short execution times.