Challenge 1 : Advanced STA - calculating setup slack
Below is Timing report for Advanced STA.
What will be setup slack in below path :
Post your answer of Exact slack in comment box with justification.
Macro Placement Guideline in ASIC Physical Design
Skew difference is more, required time logic is not adding
ReplyDeleteHow can you say skew is more ?
ReplyDeleteClock network delay is also part of report.
How sensit value impacts CRP ? It shouldn't.
ReplyDeleteIn advanced STA, to reduce pessimism compared to flat derates, sensit is applicable to CRPR as well.
DeleteCRPR is nothing but delay difference from common launch/capture path.
So sensit is expected for CRPR.
The data is arriving at 0.603, and the allowed time is 1.104. So the data have met the setup requirement and it met the requirement with a margin of 0.501. So the setup slack would be 0.501.
ReplyDeleteVarshith, if this would be regular STA report, I would be agree with you that slack would be +501ps.
DeleteBut this is advanced STA report...!
Hint: any idea why last line is strikethrough ?
0.501 is setup slack untill and unless last line strikethrough value is provided which is i guess statistical adjustment ?
DeleteYes Raja. you are correct. +501ps value is true only if last line is not there (in case of regular STA).
Deletecan you pls share your views on statistical adjustment why it is required and is it only for pocv since sigma is only in pocv ..pls correct if something wrong .
DeleteThanks all this blogs are very usefull
Why the clock uncertainty is in negative values, clock uncertainty will not be negative.
ReplyDeleteHi... consider uncertainty in this way...
DeleteUncertainty(margin) is used to make timing path pessimistic.
different EDA tools have different consideration of uncertainty polarity , but main goal is to make timing path pessimistic.
for setup analysis, to make timing path pessimistic, required time should be reduced..
if tool is considering -ve polarity, uncertainty value will be added to required time.
if tool is considering +ve polarity, uncertainty value will be subtracted from required time.
for hold, it will be reverse as more required time makes hold slack pessimistic.
Please let me know if any further clarification needed
Jignesh sir please answer statistical adjustment
ReplyDeleteHi VM,
DeleteDon't call me Sir! I am also learning along with all.
There are two methods statistical adjustment is calculated.
1. ( (0.003)² + (0.010)²)½ (this method is already explained by one member in Challenge 6.
2. Calculated unadjusted slack and Mean/sigma slack.
I will cover detailed explanation in upcoming post.
Hi VM,
DeletePlease let me know if you have further queries.
slack is met with 0.4472.
ReplyDeleteThank you Anil for sharing your views here...
DeleteCould you please share calculation how slack value arrived at 0.4472
This comment has been removed by the author.
DeleteI can share , if this is correct?
DeleteSure Anil.
DeleteYou can share the calculation.
As of now,I will not comment it's correct or not just for banifit of other members.
Here is what I understood from the report:
ReplyDeleteA setup slack of 0.501 was identified.
The analyzed data transition values are a very important factor.
The design includes information about the number of corners used.
Uncertainty values are mentioned.
CPR values are also reported.
Few points to correct,
Delete1. Setup slack is not 0.501. as this is POCV timing report, need to consider statistical adjustment.
2. Corner
Here, corner is different w.r.t POCV (it doesn't mean no. Of corner used)
I will explain this in details in upcoming posts. Until then stay tuned!