dirac16
Member level 5
I am trying to do my first digital synthesis design using design vision. I have got a question along the way. There are two methods to estimate power consumption to my knowledge: First method comes directly after the synthesis is done; while we are in the Design vision we go to the Attribute tab > Report Power... or we could type report_power > /path/my_power.txt. in dc shell to generate a power report. That is the first method. The other one I have heard relies on extracting VCD file and then converting it to SAIF. After the SAIF file is generated we could write a command to read in the SAIF file and then write another command to report the power in the dc shell. So, obviously the SAIF file includes all transitions and considers switching activity of all or many of the nets. So it should be more accurate than the former. But why is the former less accurate by the way? Doesn't it take care of all transitions and switching activities? How does it really estimate power?