Meta Leases 168MW AI Data Center From Reliance

Meta Platforms has agreed to lease an AI-enabled data center campus in Jamnagar, Gujarat, taking 168 MW of capacity with an option to scale, according to Meta's data center blog post and company press materials. The company states the site will be powered by renewable energy, cooled using desalinated seawater, and that Meta will cover the energy and water costs for the facility (Meta blog; about.fb.com). CNBC reports that Reliance Industries will build the data center and aims to deliver the facility within two years (CNBC). Meta's post also links Jamnagar to its subsea network investments, referencing Project Waterworth as part of planned connectivity for the region (Meta blog).

Public reporting frames the Jamnagar facility as a purpose-built, high-density data center intended to support modern AI workloads. Industry coverage notes the 168 MW figure is consistent with current hyperscaler AI campus sizing where sustained power density and low-latency connectivity are primary drivers (TechCrunch; DatacenterDynamics). The use of desalinated seawater for cooling and procurement of renewable energy are also common design choices for large-scale AI facilities in coastal regions where grid capacity and sustainability commitments intersect (Meta blog; CNBC).

Economic Times notes that Reliance-backed robotics startup Addverb Technologies is seeking $100 million in fresh funding, and that Christian Klein, CEO of SAP, told ET that India can take the lead in applying AI to industry, both items appear in ET's Daily Top 5 newsletter (Economic Times).

• •Delivery milestones and timelines for the Jamnagar campus, as reported by Reliance or independent reporting (CNBC).
• •Announcements of specific renewable-energy procurements or power-purchase agreements tied to the site (Meta blog; CNBC).
• •How hyperscalers configure regional connectivity for large-model training versus inference, including subsea capacity announcements linked to Project Waterworth (Meta blog).

Multiple outlets place the Meta-Reliance agreement in the context of a broader surge in AI infrastructure investment in India. TechCrunch and Bloomberg describe the deal as part of an ongoing trend of hyperscalers expanding capacity in new geographies to meet training and inference demand. Reporting recalls Meta's $5.7 billion investment in Jio Platforms in 2020 and the subsequent joint enterprise initiatives, noting this infrastructure agreement extends that commercial relationship into physical capacity (TechCrunch; Meta blog; CNBC).

For practitioners, on-premises or regionally located AI capacity affects model latency, data residency, and the economics of large-model training and deployment. Industry reporting highlights that hyperscaler-owned or leased capacity in India is being sized and designed with sustained power and networking to host large clusters commonly used for model training and large-context serving (TechCrunch; DatacenterDynamics).

Observers tracking the market should watch whether additional hyperscalers follow with comparable built-to-suit commitments in India, and how local policy incentives and energy availability shape total cost of ownership for AI infrastructure. Public reporting frames this Meta deal as one node in a broader industry move to diversify capacity beyond traditional Western hubs (TechCrunch; Bloomberg).

Mark Zuckerberg, quoted in Meta's announcement, said: "We're proud to be working with Reliance to build our first AI-enabled data center in India. This world-class facility in Jamnagar will help us scale our AI infrastructure globally while deepening our long-term investment in India's economy." (Meta blog).