L&T Secures Rs. 15,000 Crore AI Data Centre Order
India’s Larsen & Toubro (L&T) on Thursday said it has secured an order worth between ₹10,000 crore and ₹15,000 crore ($1.05 billion-$1.57 billion) from U.S.-based cloud platform Together AI to develop and host an artificial intelligence (AI) data centre powered by NVIDIA’s high-performance chips.
The order has been won by L&T’s AI infrastructure subsidiary, LTN Compute, marking the engineering and infrastructure major’s entry into the AI factory business. Following the announcement, L&T shares rose as much as 1.2percent during midday trading.
The AI data centre will be hosted by L&T’s Vyoma.AI unit at its campus in Chennai, Tamil Nadu. The facility will have a capacity of 10,000 NVIDIA B300 chips, which L&T described as the largest deployment of its kind in India.
The centre will support Together AI’s cloud platform, providing infrastructure for AI inference, fine-tuning and training workloads. NVIDIA’s B300 is among the company’s most advanced chips designed to support demanding AI workloads, particularly inference applications.
Also Read: Highlights from Reliance AGM 2026: Tech Growth, Jio IPO, Energy Future
The partnership with Together AI also underscores the increasing role of India as a location for AI computing infrastructure
Also Read: The Rise of Deep Tech in India: Why India Must Play the Long Game
For L&T, the project represents an expansion beyond its traditional engineering and infrastructure businesses into the rapidly developing AI infrastructure market. Through LTN Compute and Vyoma.AI, the company is positioning itself to participate in the development of large-scale computing facilities designed specifically for AI workloads.
Also Read: India's Deep-Tech Revolution: 10 AI Startups Driving Innovation
The partnership with Together AI also underscores the increasing role of India as a location for AI computing infrastructure. The Chennai facility is expected to provide the computing capacity required to support Together AI’s growing cloud-based AI services, including model training, fine-tuning and inference.



