Tata Communications commits $450M to build 10,000 km AI fibre network across India

Tata Communications Plans Major AI Connectivity Infrastructure

Tata Communications has committed $400 million to $450 million to construct approximately 10,000 km of AI-ready fibre across India1

1

2

2

. The Mumbai-based company, which provides data connections and cybersecurity services to businesses, outlined the ambitious plan in a stock exchange filing1

1

. This AI fibre network will connect India's critical AI data-centre hubs and represents one of the country's largest investments in AI connectivity infrastructure to date.

The project addresses a rapidly expanding market. Ganesh Lakshminarayanan, managing director and CEO of Tata Communications, stated the company believes data centre-to-data centre connectivity in India "could evolve into a market exceeding USD 1 billion over the medium term"2

2

3

3

. He emphasized that anchor customer commitments and contracted demand are driving a growing share of the expansion3

3

.

Three-Phase Rollout Targeting Key Data Center Hubs

The 10,000 km fibre network deployment will happen across three phases spanning three to five years4

4

. Phase one focuses on the Mumbai-Pune-Hyderabad-Chennai corridor, linking four of India's most significant AI and cloud infrastructure centers1

1

2

2

. Several cities along this corridor already host major cloud investments, with Microsoft's India South Central cloud region in Hyderabad joining existing regions in Pune, Chennai and Mumbai2

2

.

Later phases will extend to data center clusters on India's east coast and western regions, including Visakhapatnam, Bhubaneswar, Jamnagar and GIFT City in Gujarat1

1

3

3

. The network will connect these clusters with international cable landing stations, linking Indian AI infrastructure to global markets3

3

. Meta agreed earlier this year to lease its first AI-enabled data center in India from Reliance, to be built in Jamnagar, a location included in later phases of the Tata Communications project2

2

.

Massive Capacity Addition for AI Workloads

The AI-ready fibre infrastructure will add approximately 6 petabits per second of network capacity, dramatically expanding Tata Communications' current Indian network capacity of 1.25 petabits per second1

1

. The planned network will ultimately provide 7.2 petabytes of capacity according to company executives4

4

. This additional capacity is scheduled to arrive between the 2030 and 2036 financial years1

1

5

5

.

Rupesh Chokshi, chief product and technology officer at Tata Communications, emphasized that "the AI economy needs more than compute. It needs a network built for speed, scale and resilience"2

2

3

3

. The network is designed around ultra-low latency, high capacity and resilience to meet service-level requirements of hyperscalers, neocloud providers and enterprise customers4

4

. For the Mumbai-Chennai route, the company targets a round-trip latency of 15-16 milliseconds using multiple fibre pairs and higher-grade fibre4

4

.

Targeting Hyperscalers and Cloud Providers

Tata Communications is designing the AI Network Fabric specifically for global hyperscalers, AI platforms and enterprises2

2

3

3

. The company already has significant commitments from major cloud providers. Last year, Tata Communications agreed to build an 18,000 km fibre network for Amazon Web Services, linking its sites in Mumbai, Hyderabad and Chennai, with completion expected by the end of 20261

1

. Chokshi cited this dedicated network for Amazon Web Services and connectivity purchased by a neocloud provider across Mumbai, Chennai and Singapore as examples of growing demand4

4

.

Amazon, Microsoft, Reliance Industries and Adani Group are among firms committing roughly $300 billion to India's data centers and related facilities2

2

. The network will provide low-latency connectivity between data centres and link them to Tata Communications' global network backbone, which includes more than 500,000 km of subsea optical fibre and over 200,000 km of terrestrial fibre3

3

. The new routes will integrate with the company's existing network of around 700,000 km of undersea and land fibre1

1

.

Funding and Market Context

Tata Communications will finance the project through operating cash flows, according to company statements1

1

4

4

. The company posted 9.6% year-on-year revenue growth in the September quarter at Rs 6,688 crore, with net profit surging 86% to Rs 341 crore and free cash flows turning positive at Rs 212 crore4

4

. Its existing network capacity utilization stands at around 70%-80%5

5

.

Other Indian companies are simultaneously investing in AI infrastructure. E2E Networks announced plans to expand its AI factory capacity by nearly 59%, adding around 5.4 megawatts of capacity through NVIDIA HGX B300 systems and other Blackwell GPUs5

5

. E2E Networks' operating revenue jumped 411% to Rs 223.8 crore in Q2 FY27 from Rs 43.8 crore year-over-year, posting a consolidated net profit of Rs 72.7 crore and signing a multi-year cloud GPU contract worth approximately Rs 1,000 crore5

5

. Adani and Jabil also announced plans to build data-centre hardware in the country1

1

. These parallel investments signal India's emergence as a critical hub for global AI compute and connectivity infrastructure.