-
I ran the flow for the asap7 ibex design. After opening the OpenROAD GUI using
I further proceed to run
and then the reported timing metrics are
I have a few questions related to this procedure
|
Beta Was this translation helpful? Give feedback.
Replies: 3 comments 7 replies
-
Your two timing reports are likely using different parasitics modeling. The former uses extracted parasitics from the actual detailed routing. The latter uses parasitics estimated only from placement. This is comparing apples to oranges. For a valid comparison you need to rerun global routing, then detailed routing, then parasitic extraction.
Save the .def file and then use the conversion script in ORFS. |
Beta Was this translation helpful? Give feedback.
-
I don't think you should try to run repair design at the end of the flow. That's not a flow we support at this point. You could try increasing SETUP_SLACK_MARGIN to overfix earlier in the flow. |
Beta Was this translation helpful? Give feedback.
-
Yes. Many are documented in https://openroad-flow-scripts.readthedocs.io/en/latest/user/FlowVariables.html though this one looks to be missing. The ultimate truth is in https://github.com/The-OpenROAD-Project/OpenROAD-flow-scripts/tree/master/flow/scripts where the actual scripts that use the variables live. |
Beta Was this translation helpful? Give feedback.
Yes.
Many are documented in https://openroad-flow-scripts.readthedocs.io/en/latest/user/FlowVariables.html though this one looks to be missing.
The ultimate truth is in https://github.com/The-OpenROAD-Project/OpenROAD-flow-scripts/tree/master/flow/scripts where the actual scripts that use the variables live.