Larsen & Toubro (L&T), through its AI infrastructure businesses, has secured a major contract with U.S.-based AI cloud company Together AI to build what the companies describe as India’s largest single-cluster AI infrastructure deployment. The project will create an NVIDIA B300 AI Factory with 10,000 GPUs at Vyoma.AI’s Chennai data center campus, positioning the facility as a significant addition to India’s rapidly expanding AI compute capacity. The deployment targets large-scale inference, fine-tuning and training workloads rather than conventional enterprise computing, reflecting the infrastructure shift toward dedicated AI factories. Meanwhile, for L&T, the deal also advances its broader ambition to develop gigawatt-scale infrastructure capable of supporting increasingly compute-intensive AI workloads.

Chennai campus becomes foundation for AI Factory expansion

Vyoma.AI, an L&T company, operates the Chennai campus through its AI infrastructure subsidiary LTN Compute, giving the project an infrastructure base designed for substantial future expansion. The gigawatt-scale site includes a Phase 1 design capacity of 250 MW, while its power infrastructure has readiness for 150 MVA, creating room to scale AI deployments as demand increases. The 10,000 NVIDIA B300 GPUs will sit inside an integrated architecture that brings together hyperscale data center infrastructure, accelerated computing, high-performance networking and ultra-low-latency interconnects. In addition, high-throughput parallel storage and dedicated AI infrastructure operations will form a unified stack intended to simplify how customers deploy and scale demanding AI workloads.

Moreover, the partnership also highlights how AI cloud providers increasingly need infrastructure partners capable of delivering compute at industrial scale. Together AI plans to use the AI Factory to support its AI Native Cloud platform, covering training, fine-tuning and inference workloads for customers requiring substantial GPU capacity. “Making AI globally accessible is going to be the biggest infrastructure build-out in human history, and L&T understands that”, said Mr Vipul Ved Prakash, Co-founder & CEO, Together AI. “That’s exactly why we partnered with them to bring the scale, resilience and engineering excellence this moment demands to India.” The statement underlines the strategic role of physical infrastructure as AI cloud platforms move beyond software differentiation and compete increasingly on available compute, networking and operational scale.

L&T pushes its gigawatt AI infrastructure strategy

Consequently, for L&T, the project extends beyond a single GPU deployment and fits into a wider push to build an AI infrastructure ecosystem across India. “Artificial Intelligence is becoming foundational to every industry and AI Factories will power this transformation. Our deployment of an NVIDIA B300 AI Factory for Together AI marks a significant milestone in L&T’s Gigawatt AI Infrastructure Mission and reinforces our commitment to making India a global hub for next-generation AI infrastructure”, said Mr S N Subrahmanyan, Chairman & Managing Director, Larsen & Toubro. Furthermore, the company’s approach connects large-scale power and data center infrastructure with accelerated computing, creating a platform that can accommodate successive generations of AI hardware and increasingly dense workloads.

LTN Compute is building AI-ready digital infrastructure across India through hyperscale AI data centers, sovereign cloud platforms, AI Factory services, GPU-as-a-Service and managed AI platforms. Its target customers span governments, enterprises, cloud providers and AI innovators, giving the business a broader role in India’s emerging compute ecosystem. The Chennai deployment gives that strategy a high-profile anchor, particularly as organizations look for dedicated infrastructure for increasingly demanding AI workloads. Ultimately, the 10,000-GPU factory signals a deeper change: India’s AI infrastructure race is moving from individual GPU deployments toward large, integrated compute platforms designed around power, networking, storage and scalable operations.