AI investment has largely been associated with chipmakers and software companies. But as data center developers expand capacity, another group of businesses could benefit: the industrial equipment manufacturers and engineering firms supplying the infrastructure behind the AI boom.
$UBS Group AG(UBS)$ expects demand for AI infrastructure to accelerate in 2027, even as supply-chain constraints persist. In an October 16 research report, analyst Steven Fisher said industry experts attending Data Center World Expo in Singapore reported accelerating orders, rising multiyear capacity reservations and confidence that investment would continue beyond 2027.
The outlook puts three U.S.-listed companies in focus: $Caterpillar(CAT)$, $Cummins(CMI)$ and $Quanta(PWR)$.
📌 Key Takeaways
-
AI infrastructure demand: UBS expects data center investment to accelerate in 2027, potentially supporting industrial equipment and engineering businesses.
-
Three companies to watch: Caterpillar and Cummins are exposed to power-generation equipment demand, while Quanta Services could benefit from electrical-grid investment.
-
Power remains a bottleneck: Equipment shortages and manufacturing constraints could slow project delivery, even as orders increase.
-
Investor focus: The key question is whether growing infrastructure demand translates into sustained revenue and earnings growth at valuations that justify the risks.
🏗️ Why AI Infrastructure Is Becoming a Bigger Investment Theme
Building and operating AI data centers requires more than advanced processors. Facilities need reliable electricity, backup power systems and extensive electrical infrastructure to support increasingly demanding computing workloads.
As developers add capacity, demand can extend beyond technology suppliers to companies that manufacture power-generation equipment or build the infrastructure needed to connect data centers to the grid.
$UBS Group AG(UBS)$ also highlighted growing interest in behind-the-meter power generation, which allows data centers to generate electricity on-site rather than relying entirely on the public grid.
However, $UBS Group AG(UBS)$ expects most projects through 2028 or 2029 to remain primarily grid-connected, with traditional diesel backup systems. The eventual balance between grid connections, on-site generation and energy storage could influence which suppliers capture the greatest benefits.
The broader investment thesis is therefore not simply that AI demand will rise. It is that the physical infrastructure required to support that demand could become an increasingly important source of industrial growth.
⚡ 3 Stocks to Watch
1️⃣ Caterpillar (CAT): Power Generation Equipment
$Caterpillar(CAT)$ is widely known for its construction and mining equipment, but its power-generation business provides another potential avenue for exposure to data center investment.
As facilities require dependable electricity and backup power, demand for generators and related equipment could support Caterpillar's business. According to the UBS report, the company told analysts that demand for power-generation equipment remained exceptionally strong and that it was expanding production capacity.
The potential opportunity extends beyond the initial sale of equipment. Sustained data center construction could also support demand for related services and maintenance.
What investors should watch: Whether Caterpillar can expand production quickly enough to meet demand, and how much incremental growth its power-generation business contributes to overall earnings. Its construction-equipment business and broader industrial exposure remain important factors in its financial performance.
2️⃣ Cummins (CMI): Backup Power and Engine Technology
$Cummins(CMI)$ is another potential beneficiary of growing demand for reliable data center power.
The company manufactures engines and power-generation systems, including equipment used in standby-power applications. As data centers expand, backup systems become increasingly important because interruptions to electricity supply can disrupt critical computing operations.
The opportunity is not limited to new construction. A growing installed base of generators can also create demand for servicing and replacement equipment over time.
Nevertheless, data center demand is only one part of Cummins' business. Its performance also depends on other end markets, including commercial vehicles and industrial applications.
What investors should watch: Growth in power-generation sales, production capacity, order visibility and profitability. Investors should distinguish the potential contribution from data centers from the company's performance across its other businesses.
3️⃣ Quanta Services (PWR): The Electrical Infrastructure Connection
While Caterpillar and Cummins are more directly exposed to power-generation equipment, $Quanta(PWR)$ offers a different angle on the same theme.
Quanta provides infrastructure solutions for the electric power and other utility-related markets. As data center developers seek sufficient electricity to operate new facilities, additional investment may be required in grid connections, transmission systems and related electrical infrastructure.
This could support demand for Quanta's engineering and construction capabilities, particularly where power infrastructure needs to expand alongside new data center capacity.
The company's published second-quarter 2026 results showed total backlog of approximately $53.4 billion, providing visibility into contracted and estimated future work. However, backlog is not the same as guaranteed revenue or profit, and project timing can change.
What investors should watch: Backlog growth, project execution, labor availability, margins and the pace at which infrastructure projects convert into recognized revenue. The extent to which data center-related work contributes to overall growth also matters.
📊 Three Different Ways to Play the AI Infrastructure Theme
|
Company |
Primary exposure |
Key investor question |
|
Power-generation equipment |
Can capacity expansion keep pace with demand? |
|
|
Engines and backup power systems |
How much incremental growth will data centers contribute? |
|
|
Electrical-grid and infrastructure construction |
Can its backlog translate into profitable growth? |
The three companies are exposed to different parts of the infrastructure buildout. Caterpillar and Cummins are positioned around equipment demand, while Quanta's opportunity is more closely tied to electrical infrastructure investment.
This distinction matters because the timing of equipment orders, construction projects and revenue recognition may differ considerably.
⚠️ What Could Limit the Opportunity?
Despite the favorable demand outlook, several risks could prevent the AI infrastructure boom from translating into stronger returns for investors.
-
Supply-chain constraints: Shortages of power equipment, cooling systems and manufacturing capacity could delay construction and limit near-term deliveries.
-
Project delays or cancellations: Data center development depends on financing, access to electricity, permitting and customer demand. Announced projects do not always proceed as planned.
-
Changing power technology: Advances in power distribution, energy storage and on-site generation could change equipment requirements and the mix of suppliers benefiting from new investment.
-
Broader industrial weakness: Inflation, interest rates and slower activity in other end markets could offset some of the gains from data center demand.
-
Valuation risk: Investors may already be pricing in substantial future growth. Even strong business results may not produce attractive stock returns if expectations are too high.
For shareholders, the central issue is whether incremental AI-related demand can produce durable earnings growth, rather than merely a temporary increase in orders.
🔎 What Comes Next?
Three developments could help investors assess whether the opportunity is becoming more tangible:
-
Order growth and backlog: Continued increases in orders and project commitments would provide evidence that demand is translating into business activity.
-
Production and execution: Companies must demonstrate that they can expand capacity, manage costs and deliver projects on schedule.
-
Earnings and valuation: Future financial results will help investors determine whether AI infrastructure is becoming a meaningful growth driver and whether current share prices adequately reflect the risks.
The 2027 outlook remains a forecast, and the benefits will not necessarily arrive at the same pace for all three companies.
🧭 The Bottom Line
$UBS Group AG(UBS)$'s outlook suggests that the AI investment cycle could create opportunities well beyond semiconductor manufacturers and software providers.
$Caterpillar(CAT)$ and $Cummins(CMI)$ offer exposure to power-generation equipment, while Quanta Services provides a route into the electrical infrastructure needed to support data center expansion.
The longer-term opportunity depends on whether rising AI-related investment translates into sustained orders, profitable growth and stronger cash generation. For investors, monitoring those fundamentals — alongside valuations and execution risks — will be more informative than relying on the AI theme alone.
💬 Your Turn: Join the Discussion
As AI data center construction accelerates, which part of the infrastructure value chain do you think has the strongest long-term potential: power-generation equipment, electrical-grid construction or semiconductor technology?
Share your thoughts below. Useful comments may receive Tiger Coins! 🪙
Markets are always moving - and sometimes, the best move is knowing what works for you.
With Treasury yields, oil prices and rate expectations keeping markets on edge this week, investors are once again thinking carefully about where to position next. There’s no one-size-fits-all choice in investing — and the same goes for Tiger Merch. This month’s hot picks are in, featuring the Tiger Toiletry Bag, Universal Travel Adapter, Tiger Umbrella and more favourites chosen by fellow Tigers.
Explore the Monthly Hot Picks in Tiger Coin Mall, now 12% OFF for a limited time.
Comments
I believe electrical infrastructure offers a compelling long-term opportunity in the AI ecosystem. Chips make AI smarter, but electricity makes AI possible.
Among the three, I find $Quanta(PWR)$ particularly interesting because expanding transmission lines and grid connections addresses a bottleneck that cannot be solved simply by buying more GPUs. $Caterpillar(CAT)$ and $Cummins(CMI)$ could also benefit as data centers require reliable backup and on-site power.
However, demand does not automatically guarantee returns. On October 21, Australia's NextDC plunged after electricity upgrade delays forced it to cut its 2027 outlook, highlighting infrastructure bottlenecks.
My view: PWR offers compelling structural exposure, while CAT and CMI provide complementary power-generation exposure. The winner will be the company that converts orders into sustainable earnings and cash flow without excessive valuation.
@Capital_Insights [贱笑]
Among the three, I would keep $Quanta(PWR)$ on my watchlist for its exposure to grid expansion, while $Caterpillar(CAT)$ and $Cummins(CMI)$ also offer opportunities in power generation and backup systems. However, I would compare their valuations, earnings growth and cash flow before investing.
I remain bullish on the long-term AI theme, but I will not chase stocks simply because they are linked to AI. I prefer companies with sustainable demand, strong fundamentals and reasonable valuations, while waiting patiently for attractive entry points.
@TigerStars @Tiger_comments @TigerClub @Capital_Insights
Semiconductors remain essential to AI, but expectations and valuations are already high, while rapid technology cycles create competitive risk. Power-generation equipment should also benefit, but demand can be more dependent on individual projects and capacity additions.
Grid infrastructure looks like a broader bottleneck. AI data centres require enormous amounts of reliable electricity, and that means more substations, transmission lines, transformers and grid connections. Importantly, these investments can benefit not only AI but also electrification and general growth in power demand.
Companies such as Quanta Services therefore interest me as a “picks and shovels” approach to the AI build-out. I would still watch order backlogs, margins and free cash flow closely. The AI theme may attract investors, but ultimately sustained earnings growth has to justify the valuation.
原因有3个:
1)AI数据中心越来越耗电。 数据中心需要稳定电力,发电机和备用电源可能迎来更多需求。
2)不只靠AI赚钱。 CAT和CMI还有其他业务,可以分散对单一市场的依赖。
3)设备维修也能带来收入。 设备卖出去后,后续维修和保养也可能创造收入。
但要注意: 需求增加不代表股价一定上涨,还要看公司盈利、竞争和买入价格。
我的最终看法
如果看好AI未来10年的发展,我会重点研究 CAT和CMI,比较谁的盈利增长更强、估值更合理。
记住:不要只投资制造AI的公司,也可以研究为AI提供电力的公司。
1)电网建设:PWR AI数据中心需要大量电力,电网建设和升级可能持续受益。风险是项目延误、成本上升,以及股价估值过高。
2)发电设备:CAT、CMI 数据中心需要稳定电力和备用发电设备。如果需求持续增长,这两家公司可能受益。但它们还有其他业务,不会只靠AI赚钱。
3)半导体:NVDA AI芯片需求仍有增长机会,但科技股估值和市场波动风险也不能忽略。
我的投资看法
如果投资10年以上,我会重点研究 PWR、CAT和NVDA,比较它们的盈利增长、现金流和估值,再决定买哪一只。
记住:AI投资不只是买芯片,也可以投资支撑AI运作的电力基础设施。但好行业不代表任何价格都值得买。
我的核心逻辑是:**AI芯片决定算力上限,电力基础设施决定算力能否真正落地。**
过去市场主要关注NVDA、AMD等芯片公司,但数据中心建设面临的瓶颈正在从GPU供应逐渐扩展到电力接入、变压器、输配电设备和施工能力。
相比另外两个方向,我认为电网建设有三个优势:
① **需求来源更加多元。** 电网升级不仅服务AI数据中心,还受益于工业电气化、可再生能源接入和老旧基础设施更新。即使AI资本开支增速放缓,部分电网投资需求仍可能持续。
② **项目周期较长。** 大型输电和变电项目往往需要多年建设,有助于提高未来收入的可见性。不过,订单积压不等于已经实现的利润。
③ **竞争壁垒不仅是技术。** 工程资质、专业施工人员、项目管理能力和交付经验同样重要,产能并不能像软件服务那样迅速复制。
但我也不会忽略PWR的风险。项目延期、劳动力成本上涨、固定价格合同的利润率压力,以及市场提前反映AI增长预期,都可能影响投资回报。
我会重点跟踪三个指标:**订单积压转化率、营业利润率和自由现金流**,而不是只看新签订单金额。
CAT和CMI同样值得关注,尤其是数据中心备用发电设备需求。但从未来5年的角度看,我更倾向电网建设,因为它解决的不只是AI算力需求,而是整个电力系统的长期瓶颈。
**我的判断是:AI投资的下一阶段,市场可能不再只奖励生产芯片的公司,也会更加重视那些真正让数据中心通电、运行并持续扩容的企业。** 🐯