https://open-source-silicon.dev logo
#gf180mcu
Title
# gf180mcu
p

Philipp Gühring

02/25/2023, 10:02 PM
@Tim Edwards gf180mcuC: M5.1 says 0.28um , KLayout DRC checks for 0.28um, but magic checks for 0.36um
t

Tim Edwards

02/25/2023, 10:20 PM
Metal 5 is top level metal at 9KAngstrom thickness (0.9um), so the rule that needs to be applied is MT.1, not M5.1.
p

Philipp Gühring

02/26/2023, 1:04 PM
True, then we need to fix the KLayout DRC, which likely uncovers more issues in other projects too!
But 9K (and the nowadays necessary 11K) needs 0.44um according to the MT.1 documentation with the asterisk!
So it seems that both KLayout and magic were wrong
f

Farag Elsayed

02/28/2023, 8:25 PM
@Amro Tork Should we exclude top metal from Mn table rules and check it only in MT DRC table ? For example: 5LM --> Mn table for M1 : M4 & MT table for M5 4LM --> Mn table for M1 : M3 & MT table for M4 and so on
t

Tim Edwards

02/28/2023, 10:15 PM
That would be preferred. I assume that since the thick metal rules are more conservative, that it is not incorrect to run both decks on M5, but it certainly is wasting compute cycles. Plus, if a violation is bad enough to break both sets of rules, then the end user is going to be confronted with two different results for the same error, which will be confusing.