Adani $5B with Google to reshape India’s AI hyperscale landscape
Adani Group is preparing to invest as much as $5 billion alongside Google in a new AI data center campus in Visakhapatnam, signaling a step-change in India’s compute capacity and telecom-network demand.
Key facts: Google–Adani AI data center plan
Alphabet’s Google has outlined a multi-year plan to invest about $15 billion between 2026 and 2030 to build an AI hub in Andhra Pradesh, anchored by a gigawatt-scale data center campus in Visakhapatnam and backed by subsea connectivity and clean energy sourcing. Adani’s prospective stake would flow through AdaniConneX, the joint venture between Adani Enterprises and EdgeConneX, and sits within a broader capex program that includes substantial new debt. Google’s global push to expand data center capacity—roughly $85 billion of spend this year alone—frames India as a priority destination for AI infrastructure.
Why it matters for India’s AI, cloud, and telecom
AI training and inference are concentrating in hyperscale clusters that demand ultra-dense power, low-latency networks, and resilient clean energy, and India is racing to localize that capability to support domestic AI services, sovereign data needs, and cloud growth. For telecom operators, cloud providers, and enterprises, the Visakhapatnam project is a catalyst for metro fiber buildouts, data center interconnect upgrades, and new subsea routes, while elevating performance expectations for 5G-enabled applications and edge workloads.
What the Visakhapatnam AI data center unlocks
The planned campus shifts India from incremental megawatt additions to gigawatt-scale AI infrastructure designed for clustered accelerators and high-throughput networking.
Gigawatt-class AI design: clusters, liquid cooling, renewables
The site’s initial power envelope of roughly 1 GW places it among the largest AI-focused builds in Asia, enabling thousands of accelerator GPUs to be linked in high-radix fabrics for training and large-scale inference. Expect high-density racks that push well beyond traditional 10–20 kW norms toward liquid-cooled deployments, along with electrical topologies and microgrid controls to integrate large volumes of renewable power. Such clustering is essential to reduce model training times and to provide elastic AI capacity for India’s public cloud regions and enterprise tenants.
Network and subsea: 400/800G DCI and diverse cables
Google’s plan references a supporting subsea cable network, which is pivotal to move model checkpoints, datasets, and cross-region traffic with predictable latency and diversity. On land, the campus will require massive east-coast long-haul capacity into major metros and dense metro rings with 400G/800G optics, coherent pluggables for data center interconnect, and dark fiber for scale-out. Carrier-neutral interconnection, expanded IX fabric presence, and low-latency connectivity to cloud on-ramps will be prerequisites for AI workloads spanning regions and edges.
Partners, financing, and India’s hyperscale competition
The initiative brings together global hyperscale design, domestic infrastructure execution, and carrier-grade network integration amid intensifying competition in India’s data center market.
Roles of AdaniConneX, EdgeConneX, and Airtel
AdaniConneX combines Adani’s local infrastructure and energy capabilities with EdgeConneX’s hyperscale data center expertise, positioning the JV to deliver purpose-built AI capacity at speed. Airtel is cited as an ecosystem collaborator, signaling that carrier partners will underpin the connectivity fabric, cloud interconnects, and potentially managed services around the campus. This model mirrors how hyperscalers increasingly co-engineer power, cooling, and network stacks with local partners to accelerate time to capacity.
Financing outlook and competitive landscape
Adani has telegraphed plans to raise roughly INR 90,000 crore in debt and deploy around INR 1.5 lakh crore in capex in the next fiscal year, pointing to a multi-asset build program where AI infrastructure is a marquee pillar. Competitive pressure is rising as other Indian conglomerates and global operators scale campuses across Mumbai, Chennai, Hyderabad, and NCR, with players such as STT GDC India, Yotta, CtrlS, Nxtra, and hyperscaler-led builds vying for anchor tenants. The magnitude of Google’s commitment in Andhra Pradesh raises the bar for power availability, sustainability, and network reach in emerging coastal hubs.
Impact on telecom networks and cloud ecosystems
The Visakhapatnam hub will ripple through India’s transport, metro, and access networks, resetting requirements for bandwidth, latency, and resilience.
Metro fiber, DCI, and 400/800G optics ramp-up
Telecom operators and wholesale fiber providers should expect surging demand for diverse long-haul routes into Visakhapatnam and multi-path metro rings around the campus. Data center interconnect capacity will move quickly toward 400G and 800G waves, with coherent pluggables and ROADMs enabling flexible scale. Carriers that can bundle wavelength services, dark fiber, and SLAs aligned to AI training windows will gain share.
Edge computing and 5G integration opportunities
As AI inference moves closer to users and industrial sites, operators can extend campus connectivity into regional edge locations, pairing 5G slicing with low-latency backhaul to the AI core. For CSPs and SIs, managed private 5G, on-prem AI gateways, and secure data pipelines into the Visakhapatnam core create new offers for manufacturing, logistics, and media customers that need deterministic performance.
Execution risks: power, sustainability, supply chain
Scaling to gigawatt-class AI capacity in a coastal city demands meticulous planning across power, cooling, supply chain, and policy domains.
Power, clean energy, and advanced cooling
Gigawatt data centers hinge on firm power plus large-scale renewables, alongside grid upgrades and storage to manage intermittency. Securing clean energy at scale in Andhra Pradesh, optimizing PUE with advanced liquid cooling, and planning for heat rejection in a humid coastal climate will be central to economics and compliance with sustainability targets. Water stewardship and resilience to cyclones and flooding must be baked into site engineering.
Supply chain and policy considerations
AI accelerator availability, high-speed optics, and liquid-cooling components are global bottlenecks, so early procurement and multi-vendor strategies are critical. On the policy side, data protection rules, incentives for data center parks, and permitting for subsea landings and transmission lines will influence timelines. Carriers and vendors that align with Make in India initiatives and localize assembly where feasible can de-risk delivery.
Next steps for vendors, carriers, and enterprises
With design and partner selection accelerating, stakeholders should position offerings and capacity ahead of the build curve to capture early demand.
Guidance for network and infrastructure vendors
Prioritize portfolios around 400G/800G optics, ZR/ZR+ coherent solutions for DCI, high-capacity ROADMs, and liquid cooling kits validated for AI racks. Pre-stage inventory in India, certify with AdaniConneX and Google ecosystems, and offer energy-efficient designs that help customers hit sustainability KPIs.
Guidance for carriers and cloud partners
Map diverse long-haul routes into Visakhapatnam, secure landing station and CLS-adjacent facilities, and productize low-latency SLAs tuned for AI training cycles. Build cloud on-ramps and managed interconnects near the campus and explore joint PPAs to package green connectivity for hyperscale and enterprise tenants.
Guidance for large enterprises and AI builders
Evaluate tenancy strategies that balance sovereign data needs with global scale, pre-book interconnect capacity, and plan for hybrid AI architectures that split training and inference across the Visakhapatnam core and regional edges. Align procurement to expected accelerator availability and validate data governance under India’s current privacy regime.
Bottom line: India’s AI infrastructure scale-up
Adani’s proposed multibillion-dollar investment alongside Google elevates India into the front rank of AI infrastructure build-outs, creating immediate opportunities—and execution tests—for telecom networks, cloud ecosystems, and enterprise AI roadmaps.
Strategic takeaway: align capacity, connectivity, sustainability
The combination of gigawatt-scale compute, subsea and metro network expansion, and clean-energy integration will define competitive advantage; those who align capacity, connectivity, and sustainability fastest will set the pace in India’s AI era.
