$LITE CEO just dropped a 100x shipment forecast and the thesis is actually fire:
1. Data centers are shifting copper → photonics
Not a trend. Physics. Fiber beats copper on speed, distance, power efficiency. Infrastructure rewire is inevitable.
2. Hyperscaler demand is in the millions per order
Old optics industry shipped ~1K units/quarter and called it a win. Now $AMZN $MSFT $GOOGL $META are ordering millions per pop.
US barely supplies Chinese DCs anymore, but China still does heavy biz in the US. New US draft ban on Chinese optical modules incoming = supply crunch locked in.
3. Optics ceiling is higher than you think
Copper-to-fiber replacement alone = multi-year tailwind.
Then you add $SPCX laser comms for satellites. This goes orbital, literally.
4. Real bottleneck = Indium Phosphide (InP)
Few fabs can produce it. Optical companies have to build their own = heavy capex, high barriers, multi-year lead times to scale.
Global players who can mass-produce high-power CW lasers: $LITE $COHR $AVGO $MTSI $AAOI. Same moat structure as HBM, arguably tighter.
Photonics is the second half. You believe in light?
1. Data centers are shifting copper → photonics
Not a trend. Physics. Fiber beats copper on speed, distance, power efficiency. Infrastructure rewire is inevitable.
2. Hyperscaler demand is in the millions per order
Old optics industry shipped ~1K units/quarter and called it a win. Now $AMZN $MSFT $GOOGL $META are ordering millions per pop.
US barely supplies Chinese DCs anymore, but China still does heavy biz in the US. New US draft ban on Chinese optical modules incoming = supply crunch locked in.
3. Optics ceiling is higher than you think
Copper-to-fiber replacement alone = multi-year tailwind.
Then you add $SPCX laser comms for satellites. This goes orbital, literally.
4. Real bottleneck = Indium Phosphide (InP)
Few fabs can produce it. Optical companies have to build their own = heavy capex, high barriers, multi-year lead times to scale.
Global players who can mass-produce high-power CW lasers: $LITE $COHR $AVGO $MTSI $AAOI. Same moat structure as HBM, arguably tighter.
Photonics is the second half. You believe in light?