Why STL is Investing US$100m in an AI Data Centre Expansion

As hyperscalers scale ever larger AI campuses across the US, the pressure is growing on infrastructure providers to deliver the dense optical connectivity needed to keep vast GPU clusters communicating at speed.

That demand is now driving fresh manufacturing investment from Sterlite Technologies Ltd., which plans to invest up to US$100m in the US to expand production for AI data centre and telco customers.

STL's investment is expected to create between 400 and 500 jobs. It will also increase STL’s capacity to produce terminated optical fibre cables and other connectivity solutions for AI-ready infrastructure.

The announcement was made at the SelectUSA Investment Summit on 6 May.

The SelectUSA programme is a US federal initiative focused on attracting international business investment into the country.

It connects multinational companies with economic development organisations to support job creation and infrastructure growth.

Expanding AI data centre capacity

The move reflects growing pressure on data centre operators to scale fibre connectivity as hyperscale AI deployments continue to increase GPU density and interconnect requirements across campuses.

STL said its optical portfolio is designed to support the high-capacity links needed between AI data centres, with the company positioning its infrastructure as the core foundation for what it describes as “AI Data Highways”.

The company already has agreements in place with major US customers to support high-speed optical connectivity and data centre requirements using products including its Celesta 6912 Fibre cable.

STL recently expanded its US AI data centre offering further through the launch of its Neuralis portfolio.

Rather than focusing solely on cabling, STL is positioning itself as a vertically integrated infrastructure supplier capable of supporting the full optical connectivity chain for AI environments.

“By owning the entire value chain-from glass to data centre portfolio, we are excited to enable our customers to build the physical foundation for the AI era”, says Rahul Puri, CEO at STL.

“This investment will ensure that the infrastructure required to build a strong AI backbone behind global intelligence is scalable and reliable.”

Manufacturing and supply chain focus

The expansion highlights a broader industry shift towards regionalising supply chains for critical data centre infrastructure.

As AI infrastructure demand intensifies, operators and vendors are increasingly seeking domestic manufacturing capacity to reduce deployment delays and improve resilience across fibre and networking supply chains.

The investment signals continued momentum behind optical networking technologies as operators scale campuses to support increasingly compute-intensive AI workloads.

High-density fibre infrastructure has become a critical component in AI facility design as hyperscale operators build larger clusters of GPUs.

These require faster and more efficient interconnectivity between servers, racks and campuses.