BofA on the optionality of AI subs for $META and the potential huge opportunity ahead for a Meta Cloud unit
[Meta Cloud is not a matter of if but when imho]
Last week, $CLFD announced a massive partnership that is going completely under the radar
This is one of the main reasons why I believe why data centers are placing orders TODAY and skipping the testing phase entirely
Currently, data centers are facing a logistics and engineering issue that they are completely unprepared to solve
They are literally running out of physical room to deliver power before they run out of actual electricity
To feed power to these next gen GPUs, modern data centers have to route electricity through thick copper wires encased in heavy/bulky steel conduit pipes
Imagine trying to jam 10x more fiber optic glass into the rack space while also having to run massive metal pipes full of high voltage power to the exact same rack without being able to expand the physical space
The fiber and the power cables are literally fighting each other for limited space inside the rack, ceiling, and floors
So how does this partnership change $CLFD 's stance?
VoltServer’s technology completely removes the need for these big metal pipes, by turning electricity into safe digital pulses (known as FMP)
This allows Clearfield to run thousands of watts of power through thin data cables tucked inside the exact same lightweight cassette as their fiber optics
By combining VoltServer’s tech with Clearfield's NOVA cassette, data centers can shrink the physical space required for power infrastructure by up to 84% (not a typo)
This allows DC's to clear up huge amounts of space to deploy more racks of GPU's per sq ft which means DC's bring in more $$$
This is MASSIVE because it allows Clearfield to now sell NOVA as single "Data + Power" infrastructure product
The fact that DC's are skipping this testing phase and driving Clearfield's backlog up 39% is a sign that this product is now in high demand
This bypasses the 3 year electrical hardware shortage and lets them deploy high powered AI clusters inside their existing real estate today
So not only is Clearfield now allowing data centers to cram more GPUs into their existing floor space, but they are also slashing utility & operational bills by eliminating power waste
The economic math is simple:
Lower power bills from less energy conversion losses + more physical floor space = more GPU racks generating revenue
$CLFD 's NOVA brings in a clear ROI at scale
Consumables going brrrrrrr... $Q
When the value of the wafer goes up, the value of the underlying chemistry goes up with it
If you didn't know, semiconductor materials are priced based on the value of the final chip, not what the chemicals cost to mix
This move will now force foundries to completely abandon cheaper, legacy chemical suppliers and stick to the highest grade, lowest defect consumables only
In which Qnity’s atomic accuracy portfolio has now become a mandatory insurance policy for $TSM to protect their newly hiked margins (along with others not exclusive)
When a wafer becomes 15% more expensive, the financial loss for a manufacturing defect is magnified
Management has already sanbagged the F out of guidance and with this acting as a natural tailwind for favorable pricing going forward
The earnings beats will be glorious
and they're already thinking about a possible hike next year as well....
🏁 Final thoughts:
I spent 5 weeks grinding on the Telluride before pivoting to the MDX in 48 hours.
The "obvious" choice on paper wasn’t the real value play.
Always ask: "Who’s hungrier right now — the dealer or the captive?"
That’s usually where the best lease lives. 🦈
🚗 Just leased a fully-loaded 2026 Acura MDX Advance AWD for $799/mo.
The "cheaper" 2027 Kia Telluride Hybrid SX-P AWD I’d been chasing? $913/mo.
Same 36/12K terms. Same $0 cap reduction. First month at signing only.
The SUV that’s ~$10K MORE expensive on paper costs $114 LESS per month (pre tax, NJ/NY programs). 🧵👇
🎯 The takeaway:
Lease cost = depreciation + rent charge.
Both depend on the captive finance program more than the sticker price.
When shopping a lease:
1️⃣ Check Edmunds/forums for current MF and residuals
2️⃣ Identify all stackable incentives (loyalty, conquest, dealer cash, etc.)
3️⃣ Compare effective cap costs, not just MSRPs
4️⃣ Cross-shop multiple brands and captives
Never anchor on the sticker. 🎣