RE: IQE's photonics and compound semi conductor wafers are the future10 Jun 2026 04:28
Found this on X
If you are someone who is worried about this let me break it down for you
Today Semianalysis released a report to their institutional clients, someone people on X got ahold of it. Regardless, the timeline on CPO and 800VDC has been pushed back (allegedly).
I saw a @bubbleboi post 2 days ago talking about how $TSM is still struggling with thermal issues related to CPO. It does make sense, trying to package multiple hot ass chips alongside HBM, ELS's, and the optical engine is not an easy task. The thermal yield issue is going to be real, I do think TSM will get it figured out, hopefully soon.
Ok so the CPO volume ramp has been delayed, what now?
A few things to consider:
1. This is just a sell side report. I want to see some conformation from $NVDA, $TSM, and others to confirm the timeline and progress of CPO
2. Instead of CPO, NPO takes over to bridge the gap. Picture NPO as the less complex CPO, in terms of thermal management, wiring, and packaging. They share a large majority of the same components.
Why does it matter? Because Algos got ahold of this news and sold off everything with a photonics name attached to it. So stocks like $TSEM, $IQE, $LITE, $COHR, $AEHR got hit hard. BUT it doesn't matter
These companies are agnostic to the packaging method used. You still need an ELS in NPO. You still need FAU's, you still need a PIC, you still need memory. You need it all. So all of these companies are still highly constrained for capacity and the demand is robust.
What does take a hit, and more research to come on this is glass substrates. $LPK isn't a critical bottleneck as it once was in the CPO models. Because NPO takes the optical components and puts it close, but not attached to the ASIC substrate. The heat challenge is harder, so organic substrates can still be used for NPO.
3. This news is especially good for $AAOI, and $CRDO. $AAOI wins on volume for pluggable which will extend further into 2030 now.
$CRDO wins especially as they have footholds in both copper, and optics via Dust Photonics. Their M&A has been highly targeted and accretive. They acquired DustPhotonics for silicon photonics integration, Hyperlume for advanced MicroLED optical tech, and CoMira for link-layer IP and security. They are becoming highly vertically integrated for scale up and scale out.
4. $SIVE hmmm. With Sivers I think it does hurt the short-medium term picture since they essentially loose Ayar as a client. Ayar is so heavily focused on CPO, they don't have the ability to go towards NPO. BUT they still have $JBL which gets an upgrade via LRO volume ramp now. They also have the $GFS ecosystem which benefits from AMD, new space deals, telecom deals, and other customers who still require CW lasers.
Not exactly the thing you want to see on a Tuesday. The reality is that adapting and understanding is more important. I recommend not freaking out and being irrational. Sit with this news, see how others
Sign In
Follow the stocks
