PROJECT TITLE :

Machine learning predictive modelling high-level synthesis design space exploration

ABSTRACT:

A machine learning-based mostly predictive model design area exploration (DSE) method for high-level synthesis (HLS) is presented. The tactic creates a predictive model for a coaching set till a given error threshold is reached and then continues with the exploration using the predictive model avoiding time-consuming synthesis and simulations of latest configurations. Results show that the authors?? technique is on average 1.92 times faster than a genetic-algorithm DSE method generating comparable results, whereas it achieves higher results when constraining the DSE runtime. In comparison with a previously developed simulated annealer (SA)-based mostly method, the proposed methodology is on average two.09 faster, although again achieving comparable results.