AI Infrastructure & LLMs
Large language models, AI assistants, data centres, AI chips, cooling, optical networks and the power systems behind modern artificial intelligence.
-
Three Mile Island Is Restarting to Power Microsoft’s AI
Constellation is restarting Three Mile Island Unit 1 as the Crane Clean Energy Center under a 20-year Microsoft deal. Here is what is actually planned.
-
Why AI Data Centres Are Planning Their Own Power Plants
Long grid queues are pushing AI campuses toward microgrids and co-located generation. Learn what on-site power solves—and the risks it can shift to communities.
-
AI Data Centre Water Use: What the Numbers Really Mean
AI data centre water use depends on cooling, climate and electricity supply. Explore company efficiency claims and the local figures communities need.
-
Optical AI Data Centre Networks: Why Lasers Are Replacing Copper
Explore how optical AI data centre networks move data, why co-packaged optics matters, and the power, reliability and cost trade-offs that remain.
-
AI Data Centre Liquid Cooling: Why Future Servers Need Plumbing
AI data centre liquid cooling moves heat from dense server racks. Learn how cold plates, warm water and facility design affect power and water use.
-
AI Data Centre Energy: The 1-Gigawatt Savannah River Plan
A proposed 1-gigawatt AI data centre at Savannah River would use about 2 gigawatts of on-site energy, starting with gas and moving toward nuclear.
-
Australia Plans National AI Data-Centre Standards as Power Demand Surges
Australia’s proposed AI data centre standards turn the country’s artificial-intelligence build-out from a technology story into an infrastructure test: how quickly can electricity, water and network capacity be added without shifting the bill to households and other businesses? That challenge came into sharper focus in August 2026. The Australian Energy Market Operator (AEMO) forecast a…
-
NuScale Power: Why SMRs Are Back in the Spotlight
NuScale Power is part of a nuclear energy story that is changing quickly. For many years, nuclear power was viewed as slow, expensive, politically difficult, and exposed to long construction delays. But the rise of artificial intelligence, huge data centers, grid pressure, and the global need for reliable clean electricity has pushed nuclear energy back…
-
Why AI Data Centres May Need Nuclear Power
AI data centres need huge amounts of reliable electricity. Nuclear power, especially small modular reactors, could become a key energy source for the next generation of AI infrastructure.