Sify bets on pay-per-use AI cloud model to lower entry barriers in data centre market

CHENNAI:

Digital infrastructure firm Sify Technologies has launched a new pay-per-use pricing model for artificial intelligence  cloud services, in a move it says will help lower the cost of entry for businesses looking to deploy artificial intelligence workloads in the country.

The hourly billing model—covering hosting, power, and infrastructure—will be offered at three of Sify’s hyperscale data centre campuses in Chennai, Noida and Navi Mumbai. All three sites have been certified under NVIDIA’s DGX-Ready Data Center programme, supporting up to 130 kilowatts per rack. Sify’s Navi Mumbai facility secured certification in 2024, while Chennai and Noida received approvals more recently.

The initiative marks an effort by the Nasdaq-listed company to capitalise on India's growing role in global AI development, and to position itself as a leading infrastructure partner for GPU-intensive workloads.

“By introducing colocation pricing on an hourly basis, we aim to make it faster and easier to deploy these platforms in India,” said Sharad Agarwal, chief executive of Sify Infinit Spaces, the company’s data centre subsidiary. “Our partners can bring the latest NVIDIA GPUs to India, while Sify manages all the infrastructure support for this rapidly evolving market.”

Sify will initially support NVIDIA’s H100, H200, B200, GB200 NVL72, and GB300 NVL72 platforms—including liquid-cooled configurations—through its certified facilities. The company will also offer additional services such as global connectivity, white-glove IT infrastructure support, and managed services through its converged ICT ecosystem.

Sify’s strategy, built around flexible pricing and technical certification, could appeal to a new generation of AI startups as well as established cloud players seeking to expand their regional presence.

Still, competition remains intense. Global hyperscalers such as Amazon Web Services, Microsoft Azure, and Google Cloud continue to expand aggressively in India, with their own data centre infrastructure and AI platforms.