🚨📊 Today, the market is all red, which is a frustrating day. I want to remind everyone that by 2035, the U.S. will need 123 gigawatts of electricity to power AI data centers.

I want to take a broader perspective. Many people are focused on today's stock prices, but what's truly undervalued is electricity.

By 2035, the U.S. will need an additional 123 gigawatts (GW) of power capacity just to support #AI data centers. This isn't a vision—it's an infrastructure gap that has been repeatedly verified by multiple sources.

The question isn't "Do we need computing power?" but rather—where will the electricity come from?

What's truly scarce isn't chips, servers, or even models, but rather the power resources that have already been secured, can be connected to the grid, and have scalable expansion capabilities.

And these companies below aren't holding stories—they're holding locked-in, grid-level power:

$IREN(IREN.US)

Has secured approximately 3GW of power capacity, placing it in the top tier of its peers, directly addressing high-density computing deployment needs.

$Cipher Mining(CIFR.US)

Approximately 1.6GW, among North American mining companies, it has a relatively stable and sustainable power structure.

$Galaxy Digital(GLXY.US)

Also approximately 1.6GW, its dual layout of power and computing makes it resilient in the migration of AI and digital infrastructure.

$Hut 8 Mining(HUT.US)

Approximately 1GW, it's one of the representatives that has completed the transition from "mining company" to "computing + data center."

$Terawulf(WULF.US)

Approximately 750MW, not the largest in scale, but its power structure and site selection are AI-friendly.

$Applied Digital(APLD.US)

Approximately 600MW, a typical "undervalued power shell," with its true value to be realized when future computing demand is unleashed.

The market is trading on sentiment today, but the core of the next structural trend will undoubtedly be: Who can truly meet AI's demand for power?

When computing demand grows exponentially, power doesn't scale linearly—this is precisely the starting point of divergence.

Which company do you think can be the first to turn "power advantage" into AI cash flow?

📬 I will continue to track the key inflection points of AI data centers, power infrastructure, and computing migration, sharing the structural opportunities and risk changes I observe.