GE Vernova's 116 GW Backlog Tests AI Power Demand

GE Vernova said gas equipment backlog and reservations reached 116 GW, turning AI data-center demand into a power-equipment delivery test.

By Arkolith Newsroom3 min read
a logo-free gas turbine power plant beside high-voltage equipment before an earnings call.

GE Vernova said its Gas Power equipment backlog and slot reservation agreements grew from 100 GW to 116 GW in the second quarter, while data-center orders reached more than $5 billion year to date. The July 22 results turn the AI infrastructure story into a power-equipment delivery question.

The useful test is no longer only whether hyperscalers will spend on chips and buildings. It is whether turbine, grid and electrical-equipment suppliers can deliver enough capacity before customers' AI plans collide with years-long order books.

What changed

In its GE Vernova Q2 results release, the company reported $24.2 billion of orders, up 88% organically, and $11.1 billion of revenue, up 22%. It also said backlog grew by $13 billion sequentially from equipment and services.

The most important operating number was the gas-equipment queue. GE Vernova said Gas Power equipment backlog and slot reservations rose to 116 GW from 100 GW in the prior quarter, and that it now expects at least 125 GW under contract by the end of 2026. The related GE Vernova Q2 earnings webcast page lists the presentation, release, highlights, 10-Q and transcript resources from the July 22 event.

Photo: a close view of unmarked turbine hall equipment and high-voltage switchgear

The capacity math tightened

Chief executive Scott Strazik said GE Vernova remains on track to deliver 20 GW of annual gas-turbine output in the third quarter of 2026, with 24 GW planned for 2028 and actions under way to reach 30 GW in 2030. Against a 116 GW queue, that production path makes delivery timing central to the story.

CompressorTECH2 reported that GE Vernova expects most of its planned 2030 output to be contracted by the end of 2026 and more than half of expected 2031 production to be under contract. The trade publication also tied the 116 GW figure to the company's July 22 disclosure and to demand from utilities, data centers and other power users.

That does not mean every reserved megawatt becomes an operating plant. It means buyers are trying to secure scarce equipment years in advance. For AI data centers, that is a harder constraint than another abstract capital-spending forecast.

Why data-center readers should care

The same GE Vernova release said data-center orders in Electrification reached more than $5 billion year to date, more than double the company's 2025 total. Electrification includes grid equipment and related systems, so the data-center signal is not limited to gas turbines.

This is the second-order read from a noisy AI earnings week. Chip demand can look strong while power availability becomes the binding constraint. A customer can order servers, sign a lease and still wait on generation, interconnection, transformers or switchgear.

That puts GE Vernova beside the latest Digital Realty lease-backlog test and Prologis power-pipeline question. The common thread is that AI demand is moving into physical capacity, where timing matters as much as headline spending.

What comes next

The next evidence is whether GE Vernova converts the order book without creating new execution risk. Its release raised 2026 revenue and free-cash-flow guidance, but it also cautioned that forward-looking statements depend on supply chains, factory capacity, customer projects, regulation, permitting and grid connectivity.

The countercase is straightforward: a backlog can be large because supply is scarce, not because every AI load will arrive on time. If data-center demand slows, projects are delayed or power customers cancel reserved slots, today's queue could overstate durable revenue. For now, the 116 GW number shows that the AI buildout has moved deep into the industrial supply chain.

This article is informational only and is not investment, legal, tax or accounting advice.

#GE Vernova#AI infrastructure#Data centers#Power#Earnings