Microsoft, Singtel and Lightstorm to Build AI-Focused India-Singapore Subsea Cable
Source: The Business Times
A Microsoft-led consortium including Singtel and Singapore's Lightstorm is building a new 3,600 km subsea cable linking India, Malaysia and Singapore designed for AI and cloud workloads.

A consortium led by Microsoft and Singapore-based telecom startup Lightstorm has signed contracts to build a new submarine cable system linking India, Malaysia, and Singapore, purpose-built for AI, cloud, and hyperscale workloads. The I-2SEA cable will span 3,600 kilometres and include Singapore's Singtel and Asean Cableship, along with Tata Communications and Japan's NEC Corporation.
The cable will have landing stations in Machilipatnam in the southern Indian state of Andhra Pradesh, where both Meta and Alphabet have announced data centre investments. The consortium did not disclose the investment size but described the project as a response to surging demand for AI infrastructure capacity between Southeast Asia and India — one of the world's fastest-growing data markets. Lightstorm, an Asia-Pacific AI connectivity platform, leads the initiative alongside its major consortium partners.
This project signals intensifying competition among undersea cable builders to serve the AI data centre corridor between India and Southeast Asia. The involvement of three Singapore-based entities — Lightstorm, Singtel, and Asean Cableship — underscores the Republic's central role in regional digital infrastructure. It follows a separate S$152 million investment by Tata Communications last week to expand India-Singapore subsea cable capacity, reflecting a broader buildout of the corridor.
Why it matters for Singapore: Singapore's position as Southeast Asia's premier data centre hub depends on robust, low-latency connectivity to major markets like India. Three Singapore-linked entities participating in this cable consortium reinforces the Republic's strategic role in the AI infrastructure buildout, while boosting capacity for Singapore-based enterprises running cross-border AI workloads.