Topic
Power and compute capacity of data centres built for AI workloads, measured in megawatts or gigawatts, and the industry-wide race to expand it.
SemiAnalysis counted more than 1,000 Chinese datacenters and puts China's AI-ready capacity at 24GW, against about 56GW in the US. The figure counts halls and power, so how much compute it means depends on whether export-controlled accelerators arrive to fill them.
Reality
Infrastructure revenue grew 121% to $7.4 billion in the quarter and Oracle delivered 850MW of new datacenter capacity, but a buyer negotiating multi-year AI capacity has to price the $664 billion Oracle has already contracted.
Reality
The semiconductor unit tripled to $16.7bn and cleared the Street by $1.5bn, yet the guide for the current quarter landed 0.66% light, which is what happens when revenue arrives on someone else's gigawatt schedule.
Reality
Bloomberg puts Meta's Azure Foundry bill in the hundreds of millions a year, at trillions of tokens a week. The reason is structural, not a shortfall in capability.
Reality
Etched has first-pass silicon, 400 staff and more than $1bn in booked orders. What it does not have in public is a peak FLOPS figure, a power draw, or a third-party benchmark.
Reality