chip/block area estimation...

[deepak242003]

Hello all,


Can we share what is the general procedure we follow for area estimation of analog blocks ..??

Deepak.

 

[rfsystem]

Should it be starting from the spice (verilog) netlist?

 

[deepak242003]

yes.... with the spice and schematic both jprovided ....

 

[deepak242003]

any other suggestions to fasten the estimation process..????

 

[erikl]

any other suggestions to fasten the estimation process..????

For a very rough estimation and experiences from 0.8µm down to 0.18µm processes, I'd advice: total area = (Σ over all (W*(L+2))) * n (don't forget the dummies!)
with n=4 for a not-very-experienced layouter
n=3 for a layouter with some experience (but perhaps not very dedicated)
n=2 for a very experienced and dedicated layouter (unused areas filled with filter caps!)

Very, very rough, don't nail me! ;-)

 

[deepak242003]

any other suggestions to fasten the estimation process..????

For a very rough estimation and experiences from 0.8µm down to 0.18µm processes, I'd advice: total area = (Σ over all (W*(L+2))) * n (don't forget the dummies!)
with n=4 for a not-very-experienced layouter
n=3 for a layouter with some experience (but perhaps not very dedicated)
n=2 for a very experienced and dedicated layouter (unused areas filled with filter caps!)

Very, very rough, don't nail me! ;-)

what i think about the factor ( L+ 2 ) is including diffusion extension and spacing.... for 90nm it comes
[.28 ( extension ) + .22 (spacing) ] * 2 ( both sides ) .... gives you 1... this parameters might be more in 180 nm and above so taken as 2... .

It really helped. thx..

 

[erikl]

what i think about the factor ( L+ 2 ) is including diffusion extension and spacing.... for 90nm it comes
[.28 ( extension ) + .22 (spacing) ] * 2 ( both sides ) .... gives you 1... this parameters might be more in 180 nm and above so taken as 2... .

Right! The length addition of 2µm is a very rough assumption indeed, but it won't count a lot regarding the estimation purpose, if you think of the big uncertainty re. the multiplication factor n ! ;-)