Rio AI City: Elea's 3.2 GW Hyperscale Campus Positions Rio as Global Data Center Hub

3.2 GW Hyperscale Vision Announced at Web Summit Rio 2025

Rio AI City represents the most ambitious data center infrastructure project in Latin American history, with Elea Data Centers planning a hyperscale digital campus capable of delivering 3.2 gigawatts of computing capacity at full build. Announced by Mayor Eduardo Paes at Web Summit Rio in April 2025, the project aims to transform Rio de Janeiro into the largest data center hub in Latin America and position it among the ten largest data center hubs in the world. The initiative arrives at a moment when global demand for AI computing infrastructure is accelerating exponentially, and Brazil’s $4 billion national AI plan is creating the policy and investment framework to capture a meaningful share of that demand.

The scale of the project is extraordinary by any standard. At 3.2 GW, Rio AI City would rival the largest hyperscale campuses planned anywhere in the world, competing with facilities under development in Virginia, Texas, and Singapore. For context, the entire installed data center capacity of Latin America currently measures in the low single-digit gigawatts, meaning that Rio AI City at full build would represent a transformative addition to the region’s digital infrastructure. The project reflects a bet that Brazil’s combination of abundant renewable energy, strategic geographic positioning on submarine cable routes, and growing domestic AI demand creates the conditions for a regional data center hub that can serve both local and global markets.

Phase RJO1 is already operational, providing initial capacity that demonstrates the technical viability and market demand for the project. Phase RJO2, delivering 80 MW of additional capacity, is scheduled for completion in 2026. Subsequent phases will progressively build toward the 3.2 GW target, with the timeline dependent on market demand, power infrastructure development, and regulatory approvals. The phased approach allows Elea to match capital deployment to market absorption while maintaining the option value inherent in a campus-scale development with decades of expansion potential.

The Web Summit Rio venue for the announcement was itself significant. The global technology conference attracted thousands of executives, investors, and entrepreneurs to Rio, and the mayoral announcement ensured that Rio AI City received immediate international visibility among the technology investment community. The announcement generated coverage across global technology media, positioning Rio alongside established data center markets in a way that traditional tourism and real estate marketing could not achieve.

Strategic Rationale for Rio de Janeiro as a Data Center Hub

Rio de Janeiro’s selection as the site for a global-scale data center campus reflected several structural advantages that municipal and national authorities had been cultivating. The city sits at the nexus of submarine cable routes connecting South America to Central America, North America, Europe, and Africa. These cables, which carry the vast majority of intercontinental data traffic, provide the low-latency connectivity that hyperscale data center tenants require for applications ranging from cloud computing to AI inference to real-time financial services.

Brazil’s energy matrix provided another critical advantage. With over 70 percent of electricity generated from hydropower and growing solar and wind capacity, the country offered a renewable energy profile that data center operators valued for both cost and sustainability reasons. Rio’s climate action plan and carbon neutrality commitment by 2050 aligned with the environmental mandates that major cloud computing customers imposed on their infrastructure providers. The ability to power AI workloads with predominantly renewable energy gave Rio a competitive advantage over data center locations in countries with coal- or gas-heavy generation mixes.

The energy cost structure also favored Rio. Brazil’s abundant hydroelectric capacity provided baseload power at costs below many competing data center markets, and the growing share of solar and wind generation added price stability through diversification. For data center operations where electricity represented the largest single operating cost, the combination of renewable sourcing, competitive pricing, and supply reliability created a compelling economic case that reinforced the sustainability narrative.

The COR Operations Center expansion, with its 84 servers, nearly 10 petabytes of storage capacity, and pursuit of LEED certification, demonstrated that Rio already possessed institutional expertise in operating large-scale computing infrastructure. While the COR facility served municipal operations rather than commercial computing, its technology stack and operational practices provided a reference point that helped validate Rio’s readiness for hyperscale data center operations.

Brazil’s National AI Plan and Data Center Policy Alignment

Rio AI City benefited from a convergence of national policy initiatives designed to accelerate Brazil’s digital infrastructure development. The national AI plan, launched in 2024 with a $4 billion investment commitment, established the policy framework and public investment base for AI infrastructure development. The plan built on the Brazilian AI Strategy (EBIA) launched in 2021 and the Brazilian Strategy for Digital Transformation (2022-2026), creating a layered policy environment that addressed research, development, infrastructure, and workforce development.

The forthcoming national data center policy, scheduled for launch in May 2025, promised to add tax incentives, legal security, and sector-specific regulatory rules that would further improve the economics of large-scale data center investment. For a project of Rio AI City’s scale, the tax incentive structure could represent hundreds of millions of reais in savings over the project’s life, materially improving the investment return profile and allowing more competitive pricing for tenants.

The Brazilian startup ecosystem’s $117 billion valuation in 2025, with 21.7 percent growth, represented a domestic market for data center capacity that was expanding independently of global demand. Brazil’s 869 AI startups, with 249 funded and 3 unicorns, generated computing demand that would increasingly require local infrastructure. Companies like CloudWalk, Unico, Gupy, and BLiP needed low-latency access to GPU clusters for training and inference, demand that had historically been served by data centers in the United States but could increasingly be met locally as Rio AI City scaled.

The regulatory alignment between national policy and municipal ambition was particularly effective. Federal tax incentives reduced the capital cost, the national AI plan directed public investment toward the sector, the digital transformation strategy modernized regulatory frameworks, and Rio’s municipal government provided local permitting, infrastructure, and promotional support. This multi-level policy alignment reduced the political and regulatory risk that international data center operators typically factored into emerging market investments.

Economic Impact on Rio’s Technology Ecosystem

The economic ripple effects of Rio AI City extended far beyond the data center campus itself. The project catalyzed demand for construction workers, electrical engineers, network architects, cooling system specialists, and security professionals during the buildout phases. Once operational, the facility required permanent staff for operations, maintenance, customer engineering, and facility management. The scale of a 3.2 GW campus implied a permanent operational workforce in the thousands once fully built out.

More significantly, the presence of hyperscale data center infrastructure attracted the technology companies that consumed computing capacity. AI research laboratories, cloud service providers, financial technology firms, and enterprise software companies made location decisions partly based on proximity to data center infrastructure. Rio AI City’s development accelerated the agglomeration of technology companies in the Rio market, reinforcing the demand created by the Porto Maravalley tech hub and the existing cluster of companies including StoneCo and VTEX.

The university research pipeline added another dimension. UFRJ, ranked as the best federal university in Brazil with 91 doctorate programs and active research in engineering and computer science, produced both the talent and the intellectual property that data center proximity could commercialize. PUC-Rio’s 46 Nature Index articles and focus on science and innovation, combined with FGV’s public policy expertise, created the institutional ecosystem that could translate Rio AI City’s infrastructure into economic value.

The construction phase alone represented a multi-billion-real stimulus to Rio’s economy. Data center construction required specialized skills in electrical systems, cooling infrastructure, structural engineering, and security systems that commanded premium wages. The phased buildout toward 3.2 GW meant that construction employment would be sustained over a decade or more, providing long-term visibility into the infrastructure employment pipeline that complemented other major projects including the BRT-to-VLT conversion and the New Sambadromo District.

Competitive Positioning Against Latin American Rivals

Rio AI City’s 3.2 GW vision positioned Rio to compete not only against other Brazilian cities but against the entire Latin American data center market. Sao Paulo currently hosted the largest concentration of data center capacity in the region, with its position as Brazil’s financial capital and technology hub attracting operators including Equinix, Digital Realty, and Ascenty. However, Rio’s advantages in submarine cable connectivity, renewable energy availability, and lower real estate costs created differentiation that could attract tenants seeking alternatives to Sao Paulo’s increasingly constrained and expensive infrastructure market.

Mexico City and Santiago represented the other major Latin American data center markets, but neither had announced projects approaching Rio AI City’s scale. The project’s ambition to reach the global top ten created a competitive dynamic where scale itself became an advantage, as the largest campuses could offer lower unit costs, greater redundancy, and more flexibility to accommodate the massive computing clusters that AI training workloads required.

The comparison between Rio and Sao Paulo’s tech ecosystems extended naturally to data center infrastructure. While Sao Paulo would retain its position as the primary hub for financial services computing, Rio AI City targeted the AI and hyperscale computing segment where absolute capacity, power availability, and cooling economics mattered more than proximity to the stock exchange. This specialization allowed both cities to grow their data center markets without direct competition for the same tenants.

The global competitive landscape also worked in Rio’s favor. The explosive growth in AI computing demand had created a global shortage of data center capacity, with waiting lists of 12-24 months at major facilities in Virginia, Dublin, and Singapore. This supply constraint meant that new capacity in any viable market could attract tenants who might not have considered the location under normal market conditions. Rio AI City’s timing positioned it to capture demand displaced from constrained markets, establishing a customer base that would anchor the campus during its buildout phases.

Integration With Rio’s Smart City Infrastructure

Rio AI City’s computing infrastructure created opportunities for integration with the city’s existing smart city operations. The COR Operations Center, which processed data from 10,000 cameras, 9,000 georeferenced sensors, 5,000 traffic signal sensors, and 4,000 solid waste sensors, generated analytics workloads that could benefit from local hyperscale computing capacity. The Civitas project, with its 900 radars and 50 license plate recognition cameras for AI-powered vehicle tracking, represented exactly the type of AI inference workload that benefited from proximity to high-performance computing infrastructure.

The city’s digital governance platforms including DATA.RIO, the 1746 citizen service center, and the Rio Agora transparency platform generated data assets that AI applications could leverage. The COR.Lab innovation laboratory, which fostered research and new solutions in partnership with academic institutions and private companies, provided a channel for translating Rio AI City’s infrastructure into municipal service improvements.

The expansion of public WiFi to 5,000 access points and the deployment of IoT sensors across the metropolitan area created an urban data fabric that AI applications could analyze. From optimizing traffic signal timing to predicting maintenance needs for infrastructure to improving emergency response routing, the applications of AI to city operations were numerous and growing. Rio AI City’s local computing capacity eliminated the latency penalty of sending this data to distant data centers for processing, enabling real-time AI applications that required sub-millisecond response times.

Workforce Development and Talent Pipeline

The scale of Rio AI City created workforce requirements that extended from construction trades through advanced engineering to specialized data center operations. The phased buildout demanded electrical engineers, mechanical engineers specializing in cooling systems, network architects, cybersecurity professionals, and facilities management specialists in quantities that exceeded the current local supply. Addressing this talent gap required coordinated action between Elea, the municipal government, and educational institutions.

UFRJ’s engineering programs, PUC-Rio’s computer science department, and the growing network of technical education institutions in the metropolitan area provided the academic foundation for workforce development. However, the specialized skills required for hyperscale data center operations, including expertise in power distribution at megawatt scale, precision cooling systems, and AI infrastructure management, required supplementary training programs that the industry and government were developing jointly.

The employment multiplier for data center operations was substantial. Industry analyses suggested that each direct data center job supported 3-5 indirect jobs in construction, maintenance, security, logistics, and the technology companies that colocated at the facility. At the scale of Rio AI City’s 3.2 GW buildout, the total employment impact including direct, indirect, and induced effects could reach tens of thousands of permanent positions, representing a meaningful contribution to Rio’s employment recovery trajectory.

Investment Implications and Market Outlook

For investors evaluating Rio de Janeiro’s economic trajectory, Rio AI City represented a structural investment that would compound the city’s attractiveness over decades. The project’s impact on real estate extended beyond the immediate campus area, as technology company agglomeration driven by data center proximity would increase demand for office space, residential housing, and commercial services across the broader metropolitan area.

The project’s alignment with Brazil’s $117 billion startup ecosystem and $4 billion national AI investment provided confidence in the demand trajectory. The country’s 49 percent share of Latin American venture capital and the growing concentration of AI startups created a domestic customer base that would expand in parallel with the infrastructure buildout. International demand from cloud service providers seeking Latin American capacity added incremental demand beyond the domestic market.

The combination of Rio AI City’s hyperscale ambition, the national policy framework supporting data center development, the city’s submarine cable connectivity and renewable energy advantages, and the growing domestic technology ecosystem positioned Rio for a transformation in its economic identity. The city that had been defined by oil, media, and tourism was adding AI infrastructure as a fourth pillar that could eventually rival the others in economic significance. The 3.2 GW target represented not just a data center project but a statement of intent about Rio’s future role in the global digital economy.