customer cohort
Generator sets, turbines and related services sold for data-centre power
Not sized
AI named beside another cause: the company again credits data-centre power demand to cloud computing and generative AI together, and nothing separates AI's part, so no ballpark is built on a judgment AI share. The ceiling is Power Generation sales of , of which data centres are an unstated part.
described, no sizedisclosure: described· motive: exploratory· before LLMs: expanded· layer: facilities
Power Generation sales were , up on the year, and sales to users grew , credited primarily to data-centre gensets and turbines (claims c3 and c4). Asked whether AI and data-centre demand will last, the CEO says no customer is slowing down (claim c5). The payers are still data-centre developers and operators and the cloud companies, through dealers and unnamed, so the counterparty reads mixed. The company restarts a medium-speed gas engine platform of about GW, shipping from the fourth quarter, and lead times for gas prime power reach late 2028 and 2029. The joint cause is unchanged, so nothing is sized.
Evidence: 10 quotes, 9 figures, 4 confounds, 7 from before coverage
Large reciprocating generator sets, gas turbines and turbine-related services sold for prime and backup power at data centres, reported inside the Power Generation application of the Power & Energy segment. The company names the cause jointly: energy demand to support data-centre build-out related to cloud computing and generative AI, and the CEO adds that it is not just AI. Neither the company nor any measure separates AI's part from cloud's, nor the data-centre part of Power Generation sales from the rest of the application (industrial and distributed generation), so the Power Generation line is a ceiling on a ceiling. Most equipment is sold through independent dealers; the end customers are data-centre developers and operators and the cloud companies, none named, so the counterparty reads mixed (enterprise and cloud providers, with no split given). Multi-year agreements and backlog for prime power deliver over several years and are not revenue until delivered. The company's own capacity build for large engines and turbines is not a separate channel: no capital claim names AI, and the capacity also serves oil and gas, mining, marine and the aftermarket; the capacity remarks are kept as context claims.
Why this motive
Carried from Q1: AI is still named only beside cloud computing as the cause (claims c1 and c2), and the CEO answers a question on AI demand in terms of customers' unit requests, with no measure of AI's part (claim c5). A line credited to several causes together does not meet the offensive tell.
Before LLMs: expanded
At the anchor the company already sold reciprocating generator sets as backup power for data centres and had shipped gas turbines for prime power at some, and it named data-centre growth related to cloud computing and generative AI as the driver of Power Generation. Power Generation sales were in FY2024 against in 2023, with the increase in large reciprocating engines primarily for data centres; no figure separates data-centre sales, so the level has no baseline. The size is the whole of an activity that existed before.
“Overall strength in Power Generation, for both prime and backup power applications, continues to be driven by increasing energy demands to support data center growth related to cloud computing and generative artificial intelligence (AI).”
“Power Generation – Sales increased in large reciprocating engines, primarily data center applications. Turbines and turbine-related services increased as well.”
“Reciprocating engine demand for power generation is expected to remain strong, largely due to continued data center growth relating to Cloud Computing and Generative AI.”
“Last quarter, I mentioned we are making a multi-year capital investment in our large reciprocating engine division, including increasing capacity for both new engines and aftermarket parts. This investment will approximately double output for large engines and aftermarket parts as compared to 2023.”
“We leverage these large engines across a variety of applications, including data centers, oil and gas, large mining trucks, and distributive power generation.”
“So as you probably know, traditionally, we have provided reciprocating generator sets as backup for those data centers. But what you say is very correct. That business is changing.”
“So we have had some projects now where we've shipped gas turbines to provide prime power for data centers.”
Figures
- Power & Energy external sales, Power Generation application · 2026-CQ2
- Power & Energy external sales, Power Generation application, prior-year quarter · 2025-CQ2
- Power Generation sales, year-over-year change · 2026-CQ2
- Power Generation sales, year-over-year increase · 2026-CQ2
- Power Generation sales as a share of total sales and revenues · 2026-CQ2
- Power Generation sales to users, year-over-year growth · 2026-CQ2
- Order backlog believed to be firm, all segments · as-of 2026-06-30
- Share of backlog expected to be delivered in the next twelve months · as-of 2026-06-30
- Capacity of the restarted medium-speed gas engine platform, gigawatts (about) · as-of 2026-08-04
What else could explain it
- other: Joint cause: cloud computing beside generative AI (claims c1 and c2); in Q1 the CEO said the demand is not just AI (claim c3). Nothing separates AI's part.
- line composition: Power Generation sales of also hold non-data-centre power generation; the 10-Q credits the increase of primarily, not wholly, to data-centre applications (claim c3).
- other: Price: Power & Energy price realization was favorable and long-dated orders carry escalators, so part of the growth is price.
- other: Backlog is not revenue: the firm backlog of spans all segments, with expected to deliver in the next twelve months, and gas prime power lead times reach into 2029 (claim c9).
Quotes
“We continue to anticipate full-year growth in power generation for both Cat reciprocating engines and Solar Turbines, driven by increasing energy demand to support data center build-out related to cloud computing and generative AI.”
“We continue to anticipate growth in Power Generation for both reciprocating engines and turbines and turbine-related services, driven by increasing energy demand to support data center build-out related to cloud computing and generative Artificial Intelligence (AI).”
“Power Generation – Sales increased in large reciprocating engines and in turbines and turbine-related services, primarily in data center applications.”
“Power generation grew 72%, driven by very strong demand for large gen sets and turbines used in data center applications.”
“There is a lot of discussion around AI demand. We have constant discussions with our customers, and all I can tell you is what our discussions with them is no one is slowing down at the moment. In fact, if we can get more units out, they’re asking us to give them more units.”
“We did this not just based on data center demand. That’s obviously a big driver of it, but we have aftermarket growth objectives to take care of our growing install base on large engines and turbines. These are used, as you pointed out, also in oil and gas.”
“Now, without significant investment, we are able to restart production by leveraging our existing supply base, along with our internal capacity and capability. We plan to bring about 1.5 GW of capacity back online, and shipments are expected to begin in the fourth quarter.”
“We converted the facility in under 12 months for substantially less than building a new factory. From Wamego, we now package and ship our PGM 130 product that is popular for data center power generation.”
“When it comes to gas prime, we’re pretty extended. We’re towards the back half of 2028 and into 2029. On turbines, maybe just a little bit farther than that.”
“The order backlog increased across the three primary segments, with the largest increase in Power & Energy.”
By quarter
- Q1 2026described · described, no size · exploratory
- Q2 2026described · described, no size · exploratory