Logo
FrontierNews.ai

The AI Infrastructure Gold Rush: Why Cable Makers and Construction Firms Are Winning Bigger Than Chip Designers

While the world obsesses over artificial intelligence chips and software, a quieter but equally lucrative opportunity is unfolding in the supply chains that keep data centers running. Indian manufacturers of fiber-optic cables, cooling systems, and power equipment are experiencing windfall gains as global tech giants pour billions into AI infrastructure. Meanwhile, Irish and European construction firms are exporting their data-center building expertise across the continent, transforming their business models in the process.

This phenomenon mirrors what analysts call the "picks and shovels" trade, a reference to how selling tools during a gold rush often generates more profit than searching for gold itself. For companies like Sterlite Technologies, a fiber-optic cable manufacturer in India, the shift has been dramatic. The firm recently secured a $1.1 billion multi-year contract from a US hyperscaler, sending its stock up 400 percent this year.

Why Are Infrastructure Suppliers Outpacing Chip Makers in Profitability?

The economics are straightforward: every data center, regardless of which chips it uses or which AI models it trains, requires the same supporting infrastructure. A single AI gigafactory (a large-scale compute facility with at least 100,000 cutting-edge AI chips) demands approximately 70 megawatts of raw compute power and 120 to 150 megawatts of total facility load once cooling, networking, and support systems are included. That infrastructure must be built, installed, and maintained by someone.

Telecom equipment maker HFCL reported its strongest-ever quarter in April-June 2026, with revenue more than doubling on AI demand. Founder Mahendra Nahata explained the scale of the buildout, noting that capacity expansion has shifted from a once-every-three-to-five-years event to a daily routine. Other Indian suppliers have seen similar gains: TD Power Systems shares jumped 68 percent on expectations of rising demand from AI facilities, while MTAR Technologies benefited from demand for power units used in fuel-cell systems serving big-tech operators.

The advantage for these suppliers is structural. Unlike semiconductor manufacturers, which face intense competition and require cutting-edge fabrication plants, infrastructure suppliers operate in a market where demand is growing faster than competition can scale. Additionally, these companies often have established relationships with construction firms and hyperscalers, giving them first-mover advantages in securing contracts.

How Are European Construction Firms Capitalizing on the Data-Center Boom?

Irish construction companies have quietly become specialists in building hyperscale data centers, and they are now exporting that expertise across Europe. Dublin-based John Paul Construction, traditionally known for commercial and residential properties, made a deliberate pivot in 2020 to pursue data-center specialization. Today, approximately 40 percent of its revenue comes from the sector, with the vast bulk from European projects.

The company is currently delivering large-scale projects in the United Kingdom and Germany while monitoring other high-potential markets. Chief executive Liam Kenny stated that his firm hopes the Irish data-center market will regain sustainable momentum in the coming years. Other major Irish contractors, including Sisk and Kirby, have been building hyperscale data centers across northern Europe for over a decade, with projects in Sweden, Denmark, the Netherlands, and the Nordics.

Smaller Irish suppliers have also been pulled into the ecosystem. Clad Pad, a company founded just over two years ago, is eyeing a turnover of 10 million euros this year by providing specialized cladding protection for data centers. Chief operating officer Terence McGivern noted that while the company's focus has been largely on the Continent, this is changing. "We're in discussions with two large data centres in Dublin; it's definitely starting to liven up again," McGivern said.

Steps to Understand the Data-Center Supply Chain Opportunity

  • Identify the Bottleneck: Data centers require not just chips but also fiber-optic cables for inter-data-center networking, power distribution systems, cooling equipment, and specialized construction expertise. These components are often supplied by different companies than those making semiconductors.
  • Track Demand Signals: Monitor announcements from hyperscalers like Microsoft, Google, and Amazon about new data-center capacity. US hyperscalers have committed some $57 billion in new Indian capacity over the past year, while conglomerates Adani and Reliance have pledged $100 billion towards digital and infrastructure projects.
  • Assess Geographic Advantages: India hosts nearly a fifth of the world's data but accounts for less than five percent of global data-center capacity, leaving significant room for growth. European locations like Ireland benefit from cooler climates, reducing cooling costs, and established construction expertise.

What Challenges Could Derail This Windfall?

Despite the current boom, analysts warn that valuations have become stretched, and growing opposition to energy-guzzling data-center projects poses a real risk. Data Center Watch, a tracker run by AI intelligence firm 10a Labs, reported that at least 75 US projects worth some $130 billion were delayed or blocked in the first quarter of 2026.

In Ireland, an effective moratorium on new data-center connections to the national grid was in place from 2021 until early 2026. The rapid expansion of data centers in prior years had led to massive increases in electricity demand, prompting concerns about sustainability. The United Nations recently published a report highlighting Ireland's experience with data centers as a "cautionary tale" about runaway energy consumption.

However, the Irish government appears to be reversing course. Minister for Climate, Energy and the Environment Darragh O'Brien recently likened the data-center sector to Germany's car-manufacturing industry, signaling renewed support. In India, opposition to data-center development remains in a nascent stage, and operators have adopted technologies such as closed-loop water cooling systems to reduce pressure on local resources.

Why Europe's AI Gigafactory Plan Depends on These Supply Chains

The European Commission launched an official call for tenders on July 30, 2026, to build up to seven AI Gigafactories across the EU, with a headline investment figure of more than 30 billion euros (approximately $34.6 billion USD). However, only approximately 1 billion euros of Brussels' public share is actually locked in, with the rest depending on a long-term EU budget that has not been agreed.

Each gigafactory would be equipped with at least 100,000 cutting-edge AI chips, making a single facility roughly four times more powerful than the largest AI data centers currently operating in the EU. The power consumption alone illustrates why the energy and infrastructure question is as consequential as the funding question. European industrial electricity costs are roughly two to three times those in the United States or China, meaning a facility that might cost 100 million to 200 million euros per year to power in Europe could be operated for half or a third of that cost elsewhere.

"Every data centre needs switchgear, transformers, cables, cooling systems and grid connectivity, and Indian manufacturers are already in that supply chain," noted Prateek Nigudkar, a fund manager at Shriram AMC.

Prateek Nigudkar, Fund Manager at Shriram AMC

The infrastructure suppliers and construction firms profiting from the current boom are positioned to benefit further as Europe attempts to build its own AI capacity. However, the success of Europe's gigafactory initiative depends not just on securing funding and chips, but on having reliable partners to build, cool, and connect these massive facilities. That is where the real value lies for the companies quietly winning the AI infrastructure race.