Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

FinFET Physical Dimensions

Madner

Newbie
Newbie level 2
Joined
Aug 14, 2024
Messages
2
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
14
Which physical dimensions of FinFET can we change such as nfin, tfin, hfin, lg based on fabrication possibilities?
Please explain based on recent industrial fabrication trends.
 
Last edited:
Are you interested in the "possible" or the "practical"
(these differing by how much cash you brought with you)?

You "could" change any of the above if you accepted the
development, qualification, market-acceptance burden
yourself.

Convincing an international behemoth to do it for your
amusement, not at all likely.

I don't see a trend besides a desperate race to claim
smaller and smaller, more and more. Process technology
and lithographic equipment seems to have become
"one-off", "node" by "node". You don't just get to buy a new
stepper and run old recipes finer; everything's tossed up.
So make the changes you want and do the work, or just
(if your interest is getting a working part in the near term)
get on the bus that's going your way.
 
I just want to know the physical parameter we can change as a designer which is practically possible in fabrication in FinFET. Could you please tell me?
 
As a designer, the L (direction of current flow) is fixed
by poly etch and the fin thickness (height) by poly dep.
Not for you to touch.

If lfin is the width (across current flow) then that may be
for you, provided that the foundry did not put fixed-segment
rules on you. You'd have to read up on that, my knowledge
stops at GF 12/14. nfin is surely your call, have seen maybe
100 fins in some big devices, again check specific
foundry * flow and whether you are bound by fixed
underlayers in some semicustom environment.

Really you should just go look at commercial PDKs' PCells,
their properties are all you can touch without some
negotiation with the foundry about waivers and custom flow
(which leads to custom requalification, whigh leads to your
management saying "are you off your meds, again? ...)"

The designer is more and more constrained as tools become
less and less certain to "hold the line (width)". We're already
being made responsible for device drift (used to be a reliability
criterion), reduced service life (hours per day, shirt pocket temp
range "use model"), now you get to pick either X or Y dimension
(but not both_ for critical features and orientation, what's left
besides "too many rules, let AI do it and we'll let you know if it
fails so you can try again, still knowing nothing".

At least analog still has some runway before going completely
to sh!t. Because fins svck for analog and everybody knows it.
 

LaTeX Commands Quick-Menu:

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top