They can print money. They can’t print electricians. The AI stack has a physical layer, and it gets paid no matter who wins. Builders like $PWR, $EME, $FIX, $AGX and $STRL sit underneath the software narrative. High voltage, site work, cooling and gas plants are not optional. Every announced data center needs energized land, transformers, […]
They can print money. They can’t print electricians. Read More »
$SPX 10,000 is the conservative case. And the whole thing prices off one number. Take @DrJStrategy’s model and run it at different 10Y yields, same $440 of 2027 earnings: 3.75% 26.7x ~11,700 4.65% (today) 21.5x ~9,500 6.1% (the dot com era average) 16.4x ~7,200 Same earnings, three different markets. The 10Y picks the number. That’s
SPX 10,000 Is the Conservative Case, and It All Comes Down to One Number Read More »
POWER: $BE, $WULF, $TE. CLOUD: $NBIS, $IREN, $CRWV. MEMORY: $MU, $SNDK, $SKHY. PHOTONICS: $AAOI, $LITE, $CRDO, $SIVEF. ROBOTICS: $CCXI, $VPG, $OUST, $AMBA. DRONES: $ONDS, $UMAC, $LPTH. POWER feeds CLOUD. CLOUD runs on MEMORY. MEMORY moves on LIGHT. The robots and drones at the end are where the spend finally earns its keep. Scarcity migrates up
The AI trade is not one trade. It’s a stack Read More »
Its always about DEMAND. Have that and youre 80% there. The 1999 version. Vendors financed customers to buy their gear, the fiber got laid, and most of it never got lit. The circle broke because nobody was using what got built. That’s the only question that matters here. Not whether the money moves in a
It’s Always About Demand Read More »
TE just printed its biggest revenue quarter ever: $250M, up 41% sequentially and $40M above expectations. Production reached 935 MW, with full-year output guided to the high end. That is the part of the print the market should not ignore. Demand is scaling, execution is improving, and the company is producing more volume against a
TE’s Record Quarter Is Better Than the Headline Read More »
AST SpaceMobiles next layer of optionality is sitting inside the satellites it is already launching. Each satellite carries 10 GHz of onboard processing. Management said that AI edge compute, radar and IoT applications can be built on the existing architecture, with some opportunities potentially partner funded. That matters because it expands the revenue surface without
ASTS’s Onboard Compute Could Open a Second Revenue Stream Read More »
AST SpaceMobile’s quarter told the story in two numbers: $31.5M of revenue and a $1.3B contracted backlog. The most important part of the print was not a satellite photo. It was the financing. AST borrowed $1.15B at 1.625%, with conversion set at $149 while the stock traded in the $60s. Potential dilution was under 2%.
AST SpaceMobile’s Backlog Is the Real Signal Read More »
AST SpaceMobiles Q2 release had one sentence worth more than the whole print: a joint venture planned by the top three U.S. mobile carriers to enable space-based cellular broadband for every American. The rest of the stack matters too. ASTS ended the quarter with $1.3 billion of contracted backlog, but the carrier partnership is the
AST SpaceMobile’s Carrier JV Is the Real Q2 Signal Read More »
Nothing fundamental changed. The world is still short compute. Every efficiency gain in computing history lowered the price and raised the bill. Cheaper cycles create more cycles; Jevons has been undefeated since 1865. The forced sellers are gone. The shortage isnt. That keeps the setup constructive for the infrastructure names positioned closest to incremental demand.
The Compute Shortage Still Has Room to Run Read More »
First it was GPUs. Then memory. Now it’s PHOTONICS and demand exceeds supply until mid-2027. Every AI rack is a chain: compute memory → the optics that move the data between them. As copper hits its physical wall at 1.6T, the bottleneck moves to light. And the light layer just went scarce. The 6-inch indium
Photonics: demand exceeds supply until mid-2027 Read More »
