The goal of the Tool Option Exploration stage is to achieve the maximum QoR improvement from tool options. To do this, the flow runs three implementation jobs using tool options predicted by ML strategies, as shown in the following diagram:
QoR suggestions are carried forward from the best run in Stage 1. If later steps in Stage 1 caused QoR degradation, their suggestions are dropped. Running three ML strategy runs (when three or more strategies are available) helps smooth out QoR fluctuations from any single run.
This phase uses standard implementation runs that you can run in parallel if your compute resources allow. When run in parallel, the compile time is approximately equal to one implementation run.
When the implementation runs finish, the best run selector chooses which run to take forward. The flow behaves as follows:
- Exit if timing is met.
- Proceed to stage 3 if Last Mile Timing Closure criteria are met.
- Exit if Last Mile Timing Closure criteria are not met.