While others race to build faster GPTs and AI powered systems, India is focusing on underwater cables that move data faster…
When it comes to the AI race, everyone talks about GPUs, massive data centres and cutting-edge models. But there is an invisible layer of infrastructure that powers chatbot responses, cloud workloads and enterprise applications: the high-capacity fibre-optic networks and undersea cables.
But of late, this is gaining a lot of attention, especially in India. Technology companies, telecom operators and cloud providers are all expanding their connectivity as we move towards an AI-first future.
Moving data faster has become as much of a priority as processing it.
AI Runs on Networks, Not Just Chips
One of the main requirements of training and serving AI models is the vast amount of data that is constantly required to move between users, cloud regions and data centres. To handle the load, large AI models usually distribute the workload across multiple locations. But even when the GPUs are available, poor connectivity is the reason behind increased latency and reduced efficiency.
Broadband India Forum (BIF) president TV Ramachandran realises this problem all too well: “AI processing means crunching trillions of data units and that ability can’t just rest at one place. The results of that have to be communicated on land and overseas. The data has to go through data centres, data caches, and then it has to enable intercontinental connectivity.”
As far as he is concerned, India cannot become a global AI and data hub without expanding its international subsea cable infrastructure significantly.
New Undersea Cables Are Expanding India’s AI Capacity
Going by recent reports, India is well aware of this predicament and is answering it right away. Earlier this month, Microsoft-backed Lightstorm announced the I-2SEA submarine cable, which will connect India with Malaysia and Singapore.
The 3,600-km cable will feature landing stations in Machilipatnam (Andhra Pradesh) and Singapore, thereby creating a high-capacity route specifically designed to support AI, cloud and hyperscale workloads.
The project is a partnership involving Microsoft, Tata Communications, Singapore Telecommunications, NEC Corporation and ASEAN Cableship. This shows how AI connectivity is becoming a global collaboration instead of just being a national initiative.
According to TeleGeography data from the Reuters, India currently operates 17 active submarine cables with approximately 960 Tbps of potential capacity, while at least 10 more cable systems have already been announced.
Cloud Giants Are Building Faster Global AI Highways
It’s not just telecom companies that have been investing.
Recently, Google also announced its America-India Connect initiative. It’s a five-year USD 15 billion AI infrastructure programme that includes new fibre routes, subsea cables and AI data centres to strengthen digital connectivity between India, the United States, Africa and Asia-Pacific.
Apart from that, Google is also developing its first AI hub in Visakhapatnam. There are also plans for several international submarine cable landings that will diversify India’s global connectivity beyond traditional gateways like Mumbai and Chennai.
Tata Communications have also announced new investments in subsea fibre capacity between India and Singapore. The company is focussed on the growing demand from emerging AI hubs in both the countries.
Why This Matters for India’s AI Ambitions
India already is one of the largest internet user bases in the world coupled with its rapidly growing data centre capacity. As AI workloads increase, international connectivity is an added advantage.
Reducing latency makes for faster AI inference, smoother enterprise cloud applications and better collaboration between globally distributed AI systems. Greater cable diversity also speeds things up by reducing dependence on a handful of international routes.
India’s location on the world map also gives it a critical geographical advantage thanks to its connectivity with the Middle East, Africa, North America and the rest of Asia.
Long story short, India is building all the required physical infrastructure required to host, train and serve AI models at global scale.
The Last Word
The future of AI won’t just be determined by those with the most powerful chips. It will be equally reliant on who can move data across continents the fastest, cheapest and most reliably.
And it is clear from Google and Tata’s investments that India is definitely focussed on that front. The country is quietly building the invisible foundation beneath its AI ambitions.
And if this continues, India could very well become one of the most important AI connectivity hubs, linking Asia with the rest of the digital world.
In case you missed:
- The Hidden AI Boom Transforming India’s Tech Economy
- India Is Exporting UPI Faster Than Ever. The World Is Buying In
- How the US-China Tech Rivalry is Affecting India’s Future
- Maharashtra’s AI Push Could Become the Blueprint for India’s Next Tech Revolution
- India May Be Entering Its “Applied AI” Era
- India’s Tech Renaissance: From Follower to Global Trendsetter
- Why Indian Startups Are Building “Vertical AI” Instead of the next ChatGPT
- How Operation Sindoor Defined a New Era of Technology in Defence
- India’s 3-Hour Rule Signals a New Era for Social Media Platforms
- AI Can Write Code. What Happens to India’s Millions of Software Engineers?
