I've Got a Network and if I Apply This Technique How Much Better Can I Get It Right?

The experiments we've done so far have focused more on I've got a network and if I apply this technique how much better can I get it right. Sometimes you go oh I say 15% in latency or something right which can translate into power or frame rate depending on what you're trying to achieve. And that could mean oh I could run one more network with the same unit time because this one's going to finish sooner. So these projects because they have so many networks and there's so much complexity are really sort of a puzzle fitting experience. You know your search space is very large. This might be a way to kind of get to an idea automatically through basically wrote experimentation to solve your