Meta Establishes First Indian AI Infrastructure Hub via Reliance Partnership

Meta and Reliance Launch India’s Latest AI Compute Facility

According to TechCrunch AI, Meta and Reliance Industries have formalized an agreement to construct a 168-megawatt AI-optimized data center in Jamnagar, Gujarat. The installation represents Meta’s inaugural direct infrastructure commitment to India’s burgeoning AI ecosystem. The facility will operate on renewable electricity and employ desalinated seawater for cooling—a thermally efficient design that Meta will finance entirely. Construction completion is projected for the following two years, with possibilities for scaled expansion.

This arrangement builds on Meta’s escalating entanglement with Reliance. In 2020, Meta injected $5.7 billion into Jio Platforms, Reliance’s digital services subsidiary. That initial capital infusion evolved into a $100 million joint venture announced the prior year to commercialize enterprise AI tools targeting customers across India and international markets. The compute deal now ties infrastructure directly to that strategic partnership.

India’s Accelerating Role in Global AI Infrastructure

India has rapidly materialized as the preferred geography for multinational AI builders seeking alternative compute clusters. According to TechCrunch AI, Microsoft, Amazon, Google, OpenAI, and Uber have each disclosed recent infrastructure or cloud capacity reservations within the country. Beyond U.S. vendors, AirTrunk—backed by Blackstone—committed $30 billion toward 5 gigawatts of capacity by 2030. Domestic conglomerates Adani and Tata Consultancy Services have separately unveiled their own expansion roadmaps focused on AI workloads.

The supply expansion reflects policy coordination. New Delhi has legislated tax holidays extending through 2047 for foreign cloud operators who execute their international workloads from Indian facilities, reducing operational friction and capital costs relative to other jurisdictions.

Capacity Trajectory and Market Dynamics

The compute footprint shift is quantifiable. TechCrunch AI references government data showing India’s installed capacity rose from 375 megawatts (2020) to approximately 1.5 gigawatts (2025). Independent forecasters project continued acceleration—industry consensus targets upward of 8 gigawatts by decade’s end, propelled by cloud migration, next-generation AI training pipelines, and demand for localized data residency compliance.

Meta’s participation validates India as a node in the global compute hierarchy. The Jamnagar site will anchor both Meta’s Indian enterprise footprint and its worldwide AI infrastructure network, embedding Indian capacity into the company’s distributed training and inference architecture.

Why This Matters

Meta’s infrastructure commitment signals institutional confidence that India will sustain its regulatory advantage and cost profile for the multi-year duration of AI training cycles. For competing cloud vendors already present (Microsoft, Google, Amazon), the deal raises the compute-capacity bar and may trigger capacity announcements of their own. For Indian policymakers, success here—in terms of on-time delivery, operational reliability, and power stability—could accelerate further foreign commitments and solidify India’s position alongside established U.S., European, and Asian hubs. The real constraint going forward is grid capacity: whether India’s power infrastructure can sustain the projected 8 GW without demand-side bottlenecks or rolling constraints.