An automatic routing approach for NML circuits

Nanomagnetic Logic (NML) is an emergent computation model based on the interactions between nanomagnets. However, there are several challenges and open problems in the NML design flow, where clocking schemes and routing play an essential role. We compare the tradeoffs between NML synchronized and unsynchronized routing strategies using the A* search in this work. Both algorithms outperform previous work execution times by orders of magnitude, scaling for circuits with more than 1500 logic gates. Furthermore, we compare the synchronized circuits generated by our algorithm with ToPoliNano. For different full adders sizes, we outperform their results in terms of number of nanomagnets, absolute area, and clock zones by up to 113 × , 88 × , and 1.7 × , respectively.