HCLTech has signed a memorandum of understanding to build an AI data centre in Odisha worth roughly Rs 14,257 crore (about $1.5 billion), partnering with homegrown AI company Sarvam AI on the project announced on July 24, 2026. The facility is aimed squarely at expanding India’s sovereign AI compute capacity at a time when domestic demand for GPU infrastructure is outpacing what existing data centres can supply.
The Odisha investment places HCLTech alongside a growing list of Indian IT and cloud players racing to secure land, power and GPU supply for AI infrastructure. It comes in the same week that Utho Cloud separately announced plans to deploy 10,000 GPUs, including NVIDIA’s H200, H100, A100 and Blackwell-generation RTX PRO 6000 chips, over the next seven to eight quarters to strengthen India’s sovereign AI backbone.
Why Is India Racing to Build Sovereign AI Data Centres?
India currently depends heavily on foreign-hosted AI infrastructure and imported frontier models, a dependence that carries geopolitical risk as export-control decisions by other governments can directly affect access to compute and models used by Indian businesses. Building domestic AI data centres, like HCLTech’s Odisha facility, is designed to reduce that exposure by keeping training and inference workloads for Indian users and enterprises on Indian soil, with partners like Sarvam AI providing India-specific language and reasoning models tuned for local use cases.
What Does This Mean for Indian Businesses Using AI?
For enterprises, more domestic AI data centre capacity should eventually translate into lower latency, better data residency compliance and potentially lower compute costs as supply catches up with the surge in AI adoption across banking, retail and manufacturing. Odisha’s selection also signals a broader shift of AI infrastructure investment beyond the traditional hubs of Bengaluru, Hyderabad and Mumbai into states offering land, power and policy incentives, a pattern likely to continue as more states compete for large data centre projects.
Industry Reaction and Expert Commentary
Industry observers frame the HCLTech-Sarvam AI partnership as part of a wider “AI workshop” narrative for India, where the country moves from being primarily an IT services back office to building and hosting AI infrastructure and models directly. Analysts tracking India’s AI buildout note that both compute capacity and skilled talent remain bottlenecks, meaning large announcements like this one are an important first step but will take multiple quarters to translate into operational capacity.
What Happens Next?
The Odisha facility will move from MoU to implementation over the coming quarters, with details on construction timelines, power sourcing and GPU procurement expected to follow. Sarvam AI’s involvement suggests the centre will support both foundation model training and enterprise inference workloads once operational. Watch for follow-on announcements on financing, state government incentives and hiring as the project progresses toward groundbreaking.
Frequently Asked Questions
How much is HCLTech investing in the Odisha AI data centre?
HCLTech’s MoU covers an investment of approximately Rs 14,257 crore, or around $1.5 billion, for the AI data centre project in Odisha, announced on July 24, 2026.
Who is Sarvam AI and why is it involved?
Sarvam AI is an Indian AI company building foundation models tuned for Indian languages and use cases; it is partnering with HCLTech to help power and utilise the new Odisha facility’s compute capacity.
Why does India need more domestic AI data centres?
India currently relies heavily on foreign AI infrastructure and models, and building sovereign data centres reduces geopolitical exposure while improving latency, data residency and cost for Indian enterprises adopting AI.
Leave a comment