Why I own $IREN and $NUAI?
Because power capacity is going to be one of the defining constraints on AI data center growth over the next several years.
AI models are consuming exponentially more electricity as they scale, and the traditional grid infrastructure was never built for this level of demand. Unlike past cycles where compute chips or real estate were the limiting factors, electricity supply is now the bottleneck.
According to a recent CBRE report, most grid power capacity for existing data center developments is already committed through 2030.
That means data center developers who want to build new AI facilities are finding that they cannot secure enough grid‑connected electricity in time. They can get land and they can design buildings, but they cannot get reliable power delivered from the utility grid quickly enough to meet hyperscale AI demand.
This is where IREN position becomes strategically valuable.
The company has about 4.5 GW of power capacity already secured and connected to data center facilities. In a world where grid capacity is scarce through 2030, holding a large pool of connected and contracted power is a real economic moat.
Existing power infrastructure that is ready to serve AI compute load becomes more valuable when everyone else is struggling to find capacity. That increases the leverage IREN has when negotiating long‑term data center leases and pricing power delivery.
However, for many new projects the existing grid simply isn’t enough. Hyperscalers are being forced to look at on‑site generation alternatives such as natural gas turbines, generators, or fuel cells because they cannot wait for grid upgrades that might take years.
This is the essence of the “behind the meter” model, where instead of waiting for the utility grid to deliver power, a project builds its own generation right next to the data center and supplies electricity privately to that facility. This bypasses the grid bottleneck and gives tenants the reliable, dedicated energy they need to run massive AI workloads.
That dynamic is why I also hold a position in NUAI. The company is pursuing a development pipeline focused on AI data center campuses with on‑site power infrastructure, bypassing traditional grid constraints.
In an environment where grid power capacity is pre‑booked and hard to obtain, an ability to build behind‑the‑meter power becomes a differentiator. NUAI is positioning itself as a developer and landlord of integrated power plus compute assets, which directly addresses the fundamental bottleneck the CBRE report highlights.
In simple terms, IREN benefits from having large connected power capacity in a world where grid supply is tight, and NUAI is trying to solve the same problem from the other end by building power where the grid can’t. Both strategies address the same fundamental reality that reliable power is the key constraint limiting AI data center growth.
