10,000 PIC chips requires relatively few wafers. The 2024/2025 projected quantity of transceivers doesn’t worry me at all.
20 wafers with horrible yield can supply the 2024 SOM.
I’ve seen ALD machines that can do what LWLG needs that can run 10 wafers simultaneously.
I’d assume the “poling partner” comment from the December 2022 letter was about an equipment manufacturer who can make a similar piece of equipment that does higher volume wafer poling.
The 2026 SOM that is a few hundred thousand transceivers…yea…that’s going to need more equipment/capability than 10 wafers at a time.
I’m generalizing a bit, but you catch my drift.
I think in the next year or so, LWLG keeps all back-end activities in-house. At some point in the future, a foundry will need to do it all. Or a new custom fab dedicated to polymers. Not a billion dollar fab, but a couple $10’s of millions in CAPEX I’d guess. That’s maybe a 2025+ problem. That’s also if today’s SiPh foundries don’t take on the task, but statements from LWLG indicate they’re up for it. If I was LWLG, I’d keep the back-end magic internal as long as possible. I think that’s part of the reason why we haven’t seen a tech transfer yet and in the last year or so Lebby has puckered up for secrecy.
Bottom line…I’m not worried. I just want to see the 4x200G PIC chip because that’s the money shot.
Done rambling. Pleasant hooters.
Market Data 
Markets 
AI
