The project directory contains the SAIF output file requested in the previous timing simulation run. We use this SAIF file to further guide the power analysis algorithm.
- In the main menu bar, select .
- In the Report Power dialog box,
specify the SAIF file location in the
The SAIF file, which was requested in the simulation settings prior to running timing simulation, should appear here:
- Click OK in the Report
Power dialog box.
After the Report Power command completes, the Power tab in the results windows area displays Power Report power_2.
In the Tcl console, observe that the SAIF file is read successfully and that 100% of the design nets are matched. This assures you that the generated SAIF file is correct and matched with all design nets.
- Note the change in total power (Total On-Chip Power in the Summary view) in the power_2 report compared to the power_1 report. The total power estimated in the report generated with SAIF file data will be different than the total power estimated in the vectorless run (power_1 results).
- Examine the summary and block level (On-Chip Power) power distribution in the Summary view of the Power Report.
- Go to the Signal Rate data has been set from
simulation data the SAIF file provided.
The data is color coded to indicate activity rates read from the Simulation output file.
view in the Power Report. Note that all the
- In the Summary view
of the power_1 report
(the report generated by the vectorless analysis), click on Confidence level (the following figure).
The Confidence Level is a measurement of the accuracy and the completeness of the input data that the Report Power uses while performing power analysis.
Notice that the Confidence Level is High for the vectorless analysis because less than 25% of internal nodes are user specified for Internal Activity.
- In the Summary view
of the power_2 report
(the report generated by the analysis for which you specified a SAIF file as
input), click on Confidence level (the
Notice that the Confidence Level has increased to High, because more than 25% of internal nodes are user specified for Internal Activity.