we_are_coded.by CODE · The world, decoded
БГ

Infra · Infrastructure

Power, not chips, became AI's bottleneck, and the money knows it

$1.75 billion. That's what National Grid (a major utility) is putting into a power developer building capacity right next to AI data centers. The goal is to skip the slow wait for grid interconnection.

Source
National Grid
Published
The event
Checked on
Responsible editor
Cvetelin Ivanov
How we work
MethodCorrections

National Grid announced a $1.75 billion investment in a power developer building generation capacity right next to AI data centers. The reason is simple - waiting for grid interconnection takes years, and data centers want power now.

Fact

The facts: the model is behind-the-meter power, generation right at the consumer, bypassing transmission bottlenecks. Projects are anchored with long-term offtake agreements from big tech clients. AI's bottleneck is no longer chip supply - it's megawatt supply.

Here's the trick - the problem moved. Yesterday it was a GPU-shortage. Today it's a megawatt-shortage. A chip ships in weeks, a power plant takes years. That's why money is now chasing power. Building AI infrastructure today? Ask about megawatts first. Whoever hooks up their own generation next to the servers holds the real bottleneck. That's it.

The visual is generated code-art, no third-party images.