==============================================================
Guild: wafer.space Community
Channel: ℹ️ - Information / ⁉️-questions / Butting taps
After: 2026-08-31 11:59 p.m.
Before: 2026-10-01 12:00 a.m.
==============================================================

[2026-09-23 11:18 p.m.] bailey8889
The gf180mcu process allows tap diffusion to abut mosfet source drain. 

Thinking just about the doped silicon, this is a lateral parasitic diode - not extracted as a device and not connected.

However, I think gf180mcu adds a salicide step to the diffusion to reduce resistance. This effectively creates a low resistance layer over all diffusion (not blocked) irregardless of doping.

If this is true, then abutting tap and mosfet source/drain are shorted.

In the pdk used for chipathon 2026, the magic tech file extracts abutting tap and mosfet source/drain as shorted, but the klayout tech file does not.

Does wafer.space use a modified klayout rule that shorts the abutting diffusions?

@Leo Moser (mole99)

{Attachments}
2026-09_media/cg_amp_tia-A37AC.gds
2026-09_media/image-5A796.png


[2026-09-24 7:23 a.m.] mole99
Thanks @bailey for bringing this up.

I assume this is the same situation as on ihp, correct?
There is definitely salicide being used in the gf180mcu process. There is also a salicide block (SAB) layer.

You're refering to this line in the magic tech file regarding butted taps: https://github.com/fossi-foundation/open-pdks/blob/aa3fc215a80d32437b8cca1cb3fdee819d18c4c9/gf180mcu/magic/gf180mcu.tech#L711

As far as I can see, the KLayout connectivity setup for LVS does not consider butted taps: https://github.com/fossi-foundation/globalfoundries-pdk-libs-gf180mcu_fd_pv/blob/035e6d1f4bf0acfb60606b056e7d2159cba30fef/klayout/lvs/rule_decks/general_connections.lvs#L25
Which seems to be an oversight. Do you think it makes sense to add the simple case of a butted tap, without considering SAB for now?

By the way, are you planning to use KLayout for LVS?


[2026-09-24 8:15 a.m.] bailey8889
In magic `vss` and `I_in` are shorted, but in klayout they are not.　(old post stuck in queue).


[2026-09-24 8:46 a.m.] 246tnt
Whether butted taps extract as shorter is a bit inconsistent between magic/klayout and across PDKs.

For IHP I think they are now made to be shorted because IHP actually uses that in their standard cells ( i.e. they only connect the butted tap to power/gnd and rely on the salicide to conduct to the source ).

On sky130 I think they are not extracted as shorts ( although they actually are ) because the salicide is not great and you should never rely on it for anything.

On GF180 I'm pretty sure I've also seen them recommend to always contact both source & tap even for butted taps (although here the salicide is pretty good).

{Reactions}
👍

[2026-09-27 12:08 a.m.] mithro_
Ignore me -- but my brain completed this as "buttered taps" which kind of sounds tasty 😛

{Reactions}
😂

==============================================================
Exported 5 message(s)
==============================================================