The AI trade is getting a glow-up
The AI boom spent its first act obsessing over chips, servers, and the usual suspects with trillion-dollar market caps. Now the storyline is drifting into the plumbing: photonics, aka the stuff that moves data with light instead of electrical signals. Because apparently even artificial intelligence needs a better commute.
GraniteShares filed for a Speed of Light AI ETF last week, and the name pretty much gives the game away. It’s not promising a closet full of generic AI megacaps — it’s leaning into the optics layer of the AI stack, where lasers, transceivers, fiber, and other “sounds-like-science-fiction” parts are getting more attention.
Why investors suddenly care about light-speed data
Here’s the basic problem: AI models are getting hungrier, and the pipes connecting chips, servers, and data centers are starting to sweat. Moving all that data efficiently matters almost as much as computing it. Photonics can help by pushing more data with less heat and less power. In a world where AI infrastructure budgets are already absurd, that’s a pretty compelling pitch.
You can see why big tech is giving the theme some credibility:
- Nvidia has integrated co-packaged optics into its Blackwell platform.
- Amazon has disclosed photonic I/O in Trainium2.
- Alphabet says its Jupiter network now has more than one million optical transceiver ports.
- Microsoft has flagged optical networking as part of its long-term AI infrastructure spend.
That’s not a coincidence. That’s Wall Street sniffing out a new bottleneck and trying to package it into a ticker.
ETF land loves a new narrative
Tuttle Capital already launched the Pure Play Photonics ETF, and GraniteShares is now filing its own version of the same bet. Translation: the thematic ETF machine has moved from “AI” to “AI, but with lasers.”
The big question is whether photonics becomes a real standalone bucket or just another catchy wrapper for the same AI trade with a shinier label. Either way, the message is clear: investors are no longer just chasing who builds the brains — they’re starting to ask who builds the nervous system.
Big picture: if AI is the race, photonics may be the highway upgrade everyone ends up paying for.