“…Moreover, we found that the runtime of the two operators often dominates the overall runtime of LS optimization processes-accounts for approximately 79% of the overall runtime (see Appendix B.2.2). Thus, the runtime of the two operators acts as a bottleneck to the efficiency of LS, and inefficient LS may significantly increase the Time to Market [13], [14], [15], i.e., the overall duration for developing and commercializing new chips. Moreover, inefficient LS could significantly degrade the Quality of Results (see Section 6.4).…”