exact differnce between multipliers and fingers

hi ,

i have a basic doubt when a designer in schematic says he needs 2 multiplier and 3 fingers .......while doing the layout can we instantiate a transistor with 6 fingers..if no. what problem will arise .........

finger means how many gates a transistor has while multiple is how many times a transistor is repeated.
but generally they are the same...
in the case of 3 fingers and 2 multiples, you may layout a single transistor with 6 fingers; that is acceptable.

There is a common understanding what is the difference between "fingers" and "multiplies" for frontend schematic design.


If you specify the number of fingers you can guess that the layout would use parallel stripes of gates and shared diffusion regions for the drain and sources. Most models now take this layout depend effect of sharing (reducing) the parasitic caps of the drain/sources diodes into account in the frontend schematic simulation. Therefore the schematic allows fingers and multiples as parameters. That helps the schematic designers to get a more realistic simulation w/o layout. DC wise the difference is very small because only leakage currents are impacted by the diodes.

Actual most schematic designs deliver matching requirements by handwritten notice to the layout designer. So also with a finger+multiple parameter definition the layout designer typical choose a different number of fingers and width to get a more compact design. Only integer relations for matching groups should preserved.

If for instance a cross coupled layout is required also the multiple could be changed by the layout designer.

For matching within a matching group the width should be identical.

I recommand the simulation using M-factor size and 1-scaled size.
- ex w=2u m=5, w=10u

And u will find LOD effect about finger layout effect in BSIM lib.

Rgds.