Reliance AI Data Centre: ₹1.08 lakh cr Vizag plan (2026)

What Andhra Pradesh approved and why it matters

The Andhra Pradesh government has approved the allotment of nearly 855 acres of land to Reliance Industries Ltd for a giga-scale AI Data Center (AIDC) and a Cable Landing Station (CLS) in Vizianagaram district. The proposal is positioned as one of the largest AI and data centre investment announcements in India, with a proposed cumulative investment of ₹1,08,010 crore. The approval is part of the state’s push to build an AI and hyperscale data centre corridor around Vizag and the north coastal belt. The CLS element is notable because it is designed to connect the state directly to global internet and data traffic networks. The government’s decision also signals how state policy is increasingly being used to attract compute-led industrial investment. For investors tracking Reliance’s digital strategy, this adds a clear infrastructure anchor on the east coast.

Land allotment details under the May 2026 government order

The approval was issued under G.O. Ms No. 30 by the IT, Electronics & Communications Department on May 20, 2026. The land parcel totals 854.97 acres across Polipalli, Bhogapuram West and Bhogapuram East villages in Vizianagaram district. Andhra Pradesh has highlighted the region’s advantages such as renewable energy availability, port connectivity, subsea cable access, and large land parcels. The state has also framed Vizag and the surrounding belt as a potential hyperscale corridor for future capacity expansions. The inclusion of a dedicated cable landing station suggests the project is being planned with international connectivity as a core requirement, not an add-on. The order and land clearances provide the base layer required before campus-level engineering and utility buildouts can be accelerated.

Scale of Reliance’s broader AI and data centre capex cycle

Separately, Reliance Industries is described as gearing up for a $110 billion capex commitment to data centres and AI infrastructure over the next five to seven years, covering roughly ₹10 trillion over seven years. The scope includes multi-gigawatt data centres, an edge-compute layer integrated with Jio’s network, and nationwide AI service delivery. The services component is structured through a joint venture with an initial capital commitment of approximately ₹855 crore. This framing matters because it positions the Vizianagaram land allotment as part of a larger buildout rather than a standalone facility. It also reflects a strategy where data centre capacity, connectivity, and service delivery are planned as a single stack.

Vizag campuses and the mid-2028 operational target

The article context also references an investment of approximately $15 billion over five years from 2026 to 2030 to establish India’s first gigawatt-scale AI hub. This hub is described as comprising three data-centre campuses planned across Adavivaram, Tarluvada and Rambilli villages near Vizag, with a targeted operational date of mid-2028. In addition, a separate report reference says Reliance is set to invest ₹1.6 lakh crore (over $17 billion) in a 1.5 GW data centre cluster in Visakhapatnam, including a captive solar battery storage system. Taken together, these disclosures underline the speed and scale at which compute capacity is being planned around the Vizag region. They also reinforce that power sourcing and storage are being considered early in the investment cycle.

Incentives offered under Andhra Pradesh Data Center Policy 4.0

The Andhra Pradesh government has extended a customised package of fiscal and non-fiscal incentives under the Andhra Pradesh Data Center Policy 4.0 (2024 to 2029). The incentives approved include a 25% discount on land allotment value, 100% exemption on stamp duty and registration charges, and a ₹1 per unit power tariff subsidy for 15 years. The package also includes exemption from transmission and wheeling charges for 20 years and electricity duty exemption for 15 years. Other support measures include SGST reimbursements on construction and leasing, water tariff subsidy with long-term water supply support, and OPGW fibre access discounts and right-of-way fee waivers. The policy mix indicates the state is reducing both upfront costs and long-duration operating costs, which is critical for data centres where energy and cooling are large cost buckets.

Grid and water planning: APTRANSCO and desalination discussions

The government has directed APTRANSCO to facilitate grid infrastructure development for the campus. In parallel, APIC has been asked to explore joint ownership models for a desalination plant to support the project’s long-term water requirements. These two directives are operationally important because large data centres require stable power evacuation and sustained water availability for cooling and associated campus services. By placing these responsibilities on state agencies early, Andhra Pradesh is signalling that enabling infrastructure will be treated as part of the project’s execution plan. For large-scale compute facilities, delays often emerge from utilities and local infrastructure rather than from the data hall builds alone, making these elements relevant to timelines.

Power demand tailwinds: renewables, firm clean power, and BESS

The article context links AI and hyperscale data centres to rising electricity demand in the coming years. It cites an estimate that AI and data centres could create demand for nearly 9 GW of renewable power over the next four years, and notes Avaada is in discussions with private consumers and data centre operators for renewable solutions. Separately, a ValueQuest Investment Advisors deep-dive report says data centres, green hydrogen, and rising demand for round-the-clock clean power could add 15 to 20 GW of additional annual solar demand in India from FY29. The report also states that AI inference requires 24x7 firm power and that Solar plus BESS can deliver this at scale cost-effectively. The common thread is that compute growth is increasingly tied to firmed renewable supply, not only headline solar additions.

Competitive context: Maharashtra’s large AI and data centre MoUs

The broader investment cycle is not limited to Andhra Pradesh. Maharashtra signed an MoU worth ₹1.14 lakh crore with AM Intelligence Labs for a Green Data Centre Park and an AI Compute Hub in the Mumbai region, with a stated 500 MW capacity and deployment of 2.5 lakh high-performance AI chipsets. The plan is phased, with 200 MW scheduled for completion by 2028 and 300 MW targeted by 2030, and the project is expected to generate around 8,000 direct jobs. Maharashtra also signed MoUs worth ₹1.75 lakh crore with JW Global Group and Shyam Steel, including a ₹1,13,907 crore package for a sustainable AI campus and data centre with a 1.4 GW facility. These announcements show that states are competing for hyperscale and AI-led projects using land, power, and policy as key levers.

Market impact: what matters for investors and the ecosystem

For Reliance, the Vizianagaram approval adds a concrete project pathway to its broader AI and data centre capex narrative. For the power and renewables ecosystem, the stated 9 GW renewables demand estimate linked to AI and data centres, and the ValueQuest view of 15 to 20 GW incremental annual solar demand from FY29, frame compute as a new anchor demand segment. On economics, Deven Choksey of DRChoksey FinServ has said 60% to 70% of AI data centre investment goes into power and cooling infrastructure, highlighting why concessional tariffs and grid readiness matter. Choksey also estimates that if Reliance reaches 1 GW of data centre capacity by 2030, that alone could generate upwards of $1.5 billion in annual revenue, with EBITDA margins in the 40% to 50% range. These estimates are not company guidance, but they illustrate how scale and operating power costs can shape profitability.

Key facts snapshot

What to watch next

Execution milestones will likely hinge on grid infrastructure development, water supply planning, and the pace of campus-level construction. The state has already assigned roles to APTRANSCO and APIC, which sets the stage for utility and water decisions that can influence timelines. On the industry side, additional clarity around campus capacities, power sourcing structure, and the buildout sequence would help investors map capex phasing. Separately, the multi-state pipeline of data centre and AI compute announcements suggests that land availability, subsea connectivity, and firm clean power access will be key differentiators for where capacity finally gets built at scale.