what's the difference between "implant" & &quo

In TSMC process, "P implant" layer and "AA" layer are used together to define a PMOS, and only "AA" is used to define a NMOS.

:?: what's the purpose to use two layers to realize a active area???

:?: why couldn't only use one layer, for example "Pdiff", to define PMOS and "Ndiff" to define NMOS's active area???

:?: why PMOS and NMOS use the same "AA" layer???

Thanks!!!

what's the difference between "implant" & &amp

AA is probably active area - it just means that there is no thick field oxide (FOX) so you can make contact with the diffusion. You need AA cuts wherever there is a transistor.

As for PMOS/NMOS, are you sure you're not missing a well somewhere? In an N-well process, the NMOS are made directly in the sea of substrate, diffusing N+ over a gate as I'm sure you know.
PMOS is made by diffusing an N- well, then the gate is formed and P+ is diffused over that to make the source-drain of PMOS.

I don't have the TSMC flow here, but it sounds like you are missing a well.

2 - Sure, you can do that. I worked on one process that had two layers NACTIVE and PACTIVE that defined both the ISO cut in fox and the implant. I work on a process now that you only draw the bare minimum (like AA and gate), then the whole gds file is flipped through some clever equations to define many sub-layers like guard-ring, channel stop, buried layer, NLDD, PLDD, etc. And now you don't have to worry about spacing these when you do your layout! Maybe that's what's going on - auto generated layers...

3 - Explained above, but because their AA layer only means a cut in the ISO, and probably has nothing to do with implants.