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.

For CMOS, why are there so many metal layers?

Status
Not open for further replies.

W_Heisenberg

Full Member level 4
Joined
Feb 27, 2011
Messages
217
Helped
6
Reputation
12
Reaction score
6
Trophy points
1,298
Location
Boston
Activity points
2,575
What is the benefit for more metal layers?

And what are these metals? Iron? Or anything else?
 

A typical CMOS process uses two metal layers, made from aluminium. They are needed to provide the required signal crossings and power supply rails.
 

Are you saying one layer for GND, and the other for VDD?

If so, why is there also a situation that we have like 5-10 layers?

How so?

Thanks for your help, buddy!
 

Are you saying one layer for GND, and the other for VDD?
I didn't say this. I was talking about two metal interconnect layers. VDD and VSS are sharing it with signal interconnects. There are additional contact or via layers, but they are usually not counted as metal.

As said, it's a standard CMOS process and e.g. used by full-custom and semi-custom ASICs. Total number of process masks is 13 in this case. Newer processes will partly use more metal layers to achieve higher interconnect density.
 

Routing (interconnecting standard cells ) is done over standard cells. As technology is shrinking standard cell size is decreasing and they are packed in very small area but interconnect scaling proportion to the standard cell size. So interconnect density is increasing. It is not possible to route VLSI chips using single layer. We need multiple layer of metals to route. Traditionally aluminium was used for interconnect material but aluminium has serious problem of electromigration. Now as copper is used as interconnect material as it don't have electromigration problem and resistance is also less compared to aluminium. So copper interconnects has less RC delay, hence faster.
 
Routing (interconnecting standard cells ) is done over standard cells. As technology is shrinking standard cell size is decreasing and they are packed in very small area but interconnect scaling proportion to the standard cell size. So interconnect density is increasing. It is not possible to route VLSI chips using single layer. We need multiple layer of metals to route. Traditionally aluminium was used for interconnect material but aluminium has serious problem of electromigration. Now as copper is used as interconnect material as it don't have electromigration problem and resistance is also less compared to aluminium. So copper interconnects has less RC delay, hence faster.

Yes exactly, Also the contacts for standard cells are made of tungsten.
 

Ok, so may I say more layers for interconnection then?

Thanks, man

---------- Post added at 17:51 ---------- Previous post was at 17:50 ----------

Nice anwser!
Thanks for your help.
 

Ok, so may I say more layers for interconnection then?

Thanks, man

---------- Post added at 17:51 ---------- Previous post was at 17:50 ----------

Nice anwser!
Thanks for your help.

Yes, metal layers are used for interconnection and for providing power supply to standard cells.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top