First open source PDK where Google wasn't involved...
# general
t
First open source PDK where Google wasn't involved! https://github.com/IHP-GmbH/IHP-Open-PDK
๐ŸŒ 4
๐ŸŽ‰ 2
a
That's a really great milestone.
k
Neat, looks like they're even planning a free MPW program
p
t
They've got inductors!
๐ŸŒ 2
๐Ÿ‘ 1
"the open source PDK is not intended to be used for production at this moment."---That's a downer. What, exactly, is meant by "at this moment"?
๐Ÿ˜ฎ 1
a
@Tim Edwards This is an advanced RF process. I know multiple people who used the closed source version of it. I donโ€™t the PDK is ready for work
๐ŸŒ 1
Itโ€™s just the start
p
They have a timeline hidden in their repo too
open_PDK_gantt.png
๐Ÿ™Œ 3
m
Designs with IHP sg13s SiGe HBT BiCMOS : https://slideplayer.com/slide/11288985/
๐ŸŒ 1
t
The big question here is what the price of a shuttle run is going to be. This process has a lot of masks and a lot of processing steps. SiGe is not known for being cheap.
t
Ouch. The Caravel chip is 18.6mm^2, which comes to $145,000 (US).
k
Good value on the free shuttle then ๐Ÿ˜‰
m
@Keno Fischer currently 6.2k EUR/mm^2 - respectively 5.1k EUR/mm^2 for educational Institutions - according to https://europractice-ic.com/schedules-prices-2023/ (option MINI SIC). Please take into account: Ft/Fmax=300/500GHz. (BTW: GF charges comparable MPW prices for their 130nm SiGe BiCMOS technology)
@Tim Edwards I don't expect RiscV microcontroller cores in this technology, but smaller, predominantly analogue RF designs - typically under 3mm^2, perhaps with small state machines for control. (i.e. a ring-oscillator with few hundred GHz, HF-Filters, โ€ฆ)
A little more background knowledge about IHP's SiGe technology: Free download (approx. 110 MB) : https://www.riverpublishers.com/research_article_details.php?book_id=482
๐Ÿ‘ 1
t
@Michael Strothjohann: I only meant to be using the Caravel chip size as a reference for cost, since I have a handle on what that much area is costing us for Sky130 and GF180MCU.
๐ŸŒ 1