Table of Contents
Alessandro Lombardi, chairman of Elea Data Centers, told RCRTech that the company expects artificial intelligence demand to reshape how data center capacity is deployed
In sum – what to know:
Dual demand – AI training and inference are driving different infrastructure needs, shaping Elea’s multi-market deployment strategy.
Infrastructure constraints – Power, transmission and fiber—not demand—are now the main bottlenecks for expansion.
Capital advantage – I Squared’s backing enables earlier investment and larger-scale project execution across Brazil.
Elea Data Centers is aligning its infrastructure strategy to support both large-scale AI training and distributed inference workloads, as new backing from I Squared Capital strengthens its expansion plans in Brazil.
In an interview with RCRTech, Alessandro Lombardi, founder and chairman of Elea Data Centers, said the company expects artificial intelligence demand to reshape how data center capacity is deployed, with distinct requirements emerging across different workload types.
“Training workloads typically require very large blocks of power, high-density compute environments, and campuses with room to scale over time,” Lombardi said. “Those deployments tend to prioritize energy availability, land, and the ability to support advanced cooling.”
By contrast, inference workloads are expected to be more distributed, requiring infrastructure closer to users and enterprise demand centers. “Inference is more distributed by nature. It often benefits from being closer to population centers and enterprise demand hubs where lower latency matters,” he said.
This divergence is shaping Elea’s approach in Brazil, where it operates a nationwide platform spanning key markets such as São Paulo and Rio de Janeiro. The company is combining large-scale campuses designed for expansion with metro-based facilities aimed at supporting low-latency use cases.
“From a planning perspective, we believe the winning platforms will be able to support both models simultaneously,” Lombardi said. “Centralized campuses for large-scale compute and interconnected regional facilities for lower-latency workloads.”
The strategy builds on Elea’s existing footprint of nine interconnected data center campuses, more than 300MW of installed capacity and a pipeline exceeding 1GW under development, including large-scale projects such as its planned Rio AI City.
Lombardi also emphasized that the main constraints facing the sector are shifting away from demand and toward infrastructure availability. “Scale and access to capital are increasingly important because the biggest constraints in today’s market are no longer demand-related. They’re infrastructure-related,” he said.
These constraints include access to power, transmission infrastructure and fiber connectivity—areas that require long development timelines and significant upfront investment.
“The new ownership structure strengthens Elea’s ability to move decisively in those areas,” Lombardi added. “It enhances our capacity to invest ahead of demand, engage with utilities and strategic partners earlier in the development cycle, and pursue larger, more complex projects with confidence.”
The acquisition by I Squared Capital is expected to support these efforts, as demand for AI and cloud infrastructure continues to grow in Brazil, a market still considered underpenetrated relative to more mature regions.
With its focus on renewable energy and distributed infrastructure, Elea is positioning its platform to support both hyperscale deployments and emerging AI-driven workloads across the country.
