Reconstruct the AOCV incremental derating: cell u1/u252 has late derate 1.082, early 0.924; you apply -increment -late 0.03 and -increment -early -0.03.
From PDVerse PrimeTime STA Interview Guide · pdVerse Mentor Guide
Ten-second Interview Answer
Increments add to the existing factors, so the final late derate is 1.082 + 0.03 = 1.112 and the final early derate is 0.924 + (-0.03) = 0.894. A result below 0.0 would be clamped to 0.0.
Complete Technical Explanation
The key word is -increment . It does not replace the derate already in effect on the cell, and it does not multiply it - it is added arithmetically to it. So for u1/u252, which already carries a late derate of 1.082 and an early derate of 0.924, applying -increment -late 0.03 gives a final late factor of 1.082 + 0.03 = 1.112, and applying -increment -early -0.03 gives a final early factor of 0.924 + (-0.03) = 0.894. Note that the early increment is negative, which is what makes the early factor smaller and therefore the early path faster - the two increments push the window open in both directions by the same 0.03. There is one guard on the arithmetic: if the sum came out below 0.0, the tool clamps the resulting factor to 0.0 rather than letting a negative derate produce a negative delay. Understanding this as addition rather than substitution matters because incremental derating is how you layer a small extra guardband on top of a table-driven or library-driven derate without regenerating the underlying data - you keep the characterized factor and nudge it, and the reported factor is the sum, not the increment you typed.
In Pd / Signoff
You meet this when signoff asks for a small extra guardband on a specific block or cell class without touching the variation tables. You apply the increment, then confirm with report_ocvm or the derate section of a delay-calculation report that the factor shown is the sum you expected, not the raw increment. Getting the arithmetic right is what lets you defend how much margin you actually added.
Common Trap
Assuming -increment overrides the existing derate, so people expect to see 0.03 and 0.97 in the report instead of 1.112 and 0.894. The related error is giving the early increment as a positive number, which tightens the early side rather than loosening it and quietly removes margin instead of adding it.
Expect Next
If a chain of increments drove the early factor below zero, what value would the tool actually use in delay calculation?
Topic And Primetime Commands
Variation / AOCV Basics. report_ocvmKeep building interview depth
Continue practising STA
Continue learning free
Get a practical low-power chapter
Receive the existing “Low Power and Multivoltage Fundamentals” PDF chapter and its download link by email.
Continue practising