Minimizing switchings of the function units through binding for low power

This paper considers minimizing the switchings of the function units in high-level synthesis. Switching activities on the function units are gathered through profiling the data-flow graph of the design at hand. Several thousands of random input streams are generated and used for such profiling. A genetic algorithm is employed for creation of a wide variety of random input patterns which are critical to measuring the switching activities. The switching activities are obtained and stored in a matrix. The problem of binding the function units for low power is then formulated and solved by using the proposed method. Based on the simulations, a power saving of 8%-24%, due to reduced switchings, is indicated for the considered standard benchmarks