Similarly to the deterministic ATPG, the use with the default options of the genetic algorithm based ATPG means that we should give only the name of the design to the ATPG:
 

The genetic ATPG has several options the tuning of which can help to improve the test quality. They are -mutation_rate1, -mutation_rate2, -popul_size, -timeout, and -max_generations (refer to the TT manual for an explanation of each option's function). Here we use the -popul_size to make the test shorter:

prompt> genetic -popul_size 64 c17 Genetic ATPG	*the population here is two times bigger (64)* *than the default one (32)*
Reading SSBDD-model file c17.agm... OK Allocating test patterns... OK Generating... OK Time, used by process: 0.170000 Allocating test patterns... OK Writing test patterns file c17.tst... OK Tested 22 of 22 faults Fault coverage: 100.000000 4 vectors generated

You can also try to play with -mutation_rate1 and -mutation_rate2 options. Probably they can help to achieve a better test generation time.

Last update: 28 July, 2004