Infrastructure

Primary audience: Supply-chain vendors. Paired Signal dashboard: Infrastructure Monitor.

Top story

Alphabet spent $51 billion on AI infrastructure in one quarter — and it still isn't enough

Google Cloud revenue more than doubled in Q3, but Alphabet is now renting outside compute capacity to keep up with demand, even as it warns 2027 spending will climb again.

Published today · Read the full story →

Also today

SpaceX closes $40 billion in debt to buy Nvidia chips, then Musk predicts a neocloud bloodbath

Elon Musk says SpaceX locked in $40 billion in callable debt for GPUs arriving in 2027, even as he warns smaller AI infrastructure firms are running out of credit.

Published 1 day ago · Read the full story →

Field Guide

Infrastructure terms, explained

Behind-the-meter power. Electricity generated on-site for a data center's own use, rather than drawn from the public grid. The appeal is speed: building your own gas-fired plant next to the server halls lets a developer skip the years-long wait for a grid interconnection permit. The catch is that the generation equipment has its own supply chain — which is exactly why gas-turbine delivery delays, not grid politics, are now the binding constraint on projects like Anthropic's "Project Longhorn" campus near Austin.

Interconnection queue. The wait list a utility or grid operator uses to study and approve any new large electricity user before it can actually draw power — originally built for a trickle of applications a year, now overwhelmed by hundreds of gigawatts of proposed AI data centers. Multi-year queues are the single biggest reason developers have turned to behind-the-meter generation instead, trading one bottleneck (grid approval) for another (turbine manufacturing capacity).

Simple-cycle gas turbine. A gas turbine that generates electricity directly from burning fuel, without the second steam-turbine stage that makes a "combined-cycle" plant more efficient but slower to build. Data center developers favor the simple-cycle version because it's faster to install, not because it's cheaper to run — and even that speed advantage hasn't been enough to outrun the backlog: manufacturers including GE Vernova and Siemens Energy are already quoting delivery slots into 2029 and beyond.

Large-load tariff. A special electricity rate structure a utility designs specifically for an unusually large single customer — a data center campus, for instance — rather than applying standard residential or commercial rates. The design question that makes these politically contentious is who actually pays for the new grid capacity the large customer requires: NiSource's deal to power a Microsoft data center in Indiana was structured explicitly so existing ratepayers come out ahead rather than subsidizing the buildout, a template other utilities facing similar backlash may look to copy.

Solid-state transformer. A power-conversion unit built from semiconductor modules instead of the copper windings and mineral oil inside a conventional transformer, converting utility power directly to the DC voltages a data center's racks use. The pitch is less about raw efficiency than lead time and resilience: conventional transformers now take three to four years to deliver in parts of the US, a delay that increasingly decides when a data center can actually open, while a solid-state unit's swappable modules can reportedly be replaced in minutes rather than weeks.

Vertical GaN transistor. A gallium-nitride power transistor built in a 3D "vertical" structure, so current flows through the chip's full thickness instead of across its surface the way today's "lateral" GaN transistors work — the design bet behind Vertical Semiconductor's federal CHIPS Act award, aimed at packing more current-handling capacity into the same chip area as power delivery, not raw compute, becomes the binding constraint inside dense AI server racks.

Bring Your Own Capacity (BYOC). A deal structure in which a data-center operator funds or contracts outside power capacity — batteries, smart thermostats, demand-response assets — rather than waiting on new generation or transmission, letting it add load to a strained grid without a multi-year interconnection queue.

Remaining Performance Obligation (RPO). Contracted future revenue a company hasn't recognized yet -- the accounting line that shows up as a cloud or compute "backlog." A large, fast-growing RPO signals strong pre-sold demand, but it also means a company has promised capacity it may not have finished building yet, which is why backlog growth and physical buildout speed are tracked as two separate, sometimes diverging, numbers.

Browse the full Field Guide →

More in Infrastructure

FERC Blocked Mandatory 'Bring Your Own Power' for Data Centers. Google Signed the Voluntary Version Anyway, at 10x the Size.

Days after federal regulators rejected a plan to force data centers to supply their own grid capacity, Google quietly expanded a voluntary version of the same idea tenfold — a bet that batteries and thermostats can keep pace with its AI buildout in the country's most strained grid.

Published 2 days ago

SK Group Wants $350 Billion to Turn South Korea Into an AI Data Center Hub

SK Hyper's chief executive says the company needs outside capital just for the first 5 gigawatts of a 15-gigawatt buildout plan, a pitch landing days after a Korean AI infrastructure IPO cratered 27% on its debut.

Published 3 days ago

Washington Takes an Equity Stake in a Two-Year-Old MIT Chip Startup to Fix AI's Power Problem

A $73 million letter of intent for an unproven gallium-nitride transistor maker shows the Commerce Department extending its CHIPS Act equity-stake strategy from fabs to speculative power electronics.

Published 3 days ago

Nvidia Wants to Ship Entire AI Data Centers in a Box

The company's new DSX Pod promises to cut data center build time from six months to two weeks, but only for sites that already have power, cooling and network hookups waiting.

Published 6 days ago

Vantage Bets on Startup's Solid-State Transformers to Beat Grid Bottleneck

A 1GW deal with Heron Power is the biggest commercial validation yet that chip-based transformers, not just gas turbines, could ease the grid gear shortage stalling AI buildouts.

Published 1 week ago

Microsoft Walks Away From Nscale's Flagship UK Site Over Grid Delays, Clouding a $35 Billion IPO

Microsoft has pulled out as anchor tenant at Nscale's Loughton data center after grid officials said power might not arrive until the 2030s — a blow that lands weeks before Nscale's NYSE listing push and underscores how UK interconnection queues, not chip supply, are now the binding constraint on AI buildout.

Published 1 week ago

Nscale Is Pitching Anthropic a Second $45 Billion Deal Before Financing the First

Bloomberg reports the Nvidia-backed cloud developer wants to lock in another six-year, 460-megawatt compute agreement even as its own IPO filing admits it hasn't secured financing for the GPUs behind last August's $45 billion deal, and Microsoft just walked away from a different Nscale site over grid delays.

Published 1 week ago

A gigawatt of generators: data centers are learning to skip the grid queue entirely

A Calgary equipment maker just booked over 1 GW of on-site gas generators for an unnamed data center, pushing its behind-the-meter backlog past 1.75 GW. The bigger story is what's happening at the sites already running this way.

Published 1 week ago

Amazon to fund 960MW of untested small reactors at a dead Indiana coal plant

Amazon, AEP and X-energy are exploring a 12-reactor nuclear project on the site of a demolished coal plant in Sullivan County, Indiana — the second and largest Amazon-X-energy deal yet, structured so Amazon alone, not AEP's other customers, eats the risk if the unproven reactor design runs over budget or behind schedule.

Published 1 week ago

Browse every story, including older Infrastructure coverage →