Hiranandani Group-promoted Yotta Data Services is playing a key role in advancing India’s AI ecosystem with its cloud and data centre solutions, fast-tracked by Nvidia’s advanced GPUs (graphics processing unit) and AI technologies. The Mumbai-based firm has partnered with AI startup Sarvam to build Sarvam 1, India’s first open-source foundational model. Together, they aim to accelerate AI adoption across industries in India.
Trained entirely on Yotta’s Shakti Cloud infrastructure, Sarvam 1 benefits from superior performance, scalability, and reliability. Sarvam has also launched AI agents, currently available in 10 Indian languages, that are designed to optimise various processes, including customer support, feedback collection, and employee engagement. “Yotta’s scalable and compliant GPU infrastructure has optimised our AI workflows, reducing costs and accelerating innovation,” said Vivek Raghavan, co-founder, Sarvam AI.
Similarly, RenderNet AI is a powerful tool for generating images and videos, providing good control over character design, composition, and style. Shakti Cloud enabled Rendernet to cut the costs of running inferencing workloads by efficiently managing GPU resources and scaling as needed. Yotta’s infrastructure provided Rendernet with enhanced uptime and availability, ensuring their real-time rendering processes operated without disruption. “Yotta’s infrastructure optimised our inferencing costs and ensured high uptime, allowing us to maintain top tier rendering performance,” said Bhagaban Behera, CEO & co-founder, RenderNet AI.
At Yotta, advancing India’s AI journey starts with building the country’s most powerful and sovereign AI cloud — Shakti Cloud, said Sunil Gupta, the co-founder, CEO & MD. “Shakti Cloud is India’s fastest AI cloud platform, hosting the largest concentration of Nvidia H100 GPUs in the country. It is purpose-built to empower enterprises, researchers, and startups with the raw computing power needed to train, fine-tune, and deploy LLMs and other AI workloads at scale,” he pointed out.
As AI-driven innovation gains momentum in India, startups require access to advanced IT infrastructure to build and scale breakthrough solutions. To address this, Yotta Data Services is emerging as a key player; through its data centres and Shakti Cloud, the company offers startups affordable, scalable access to advanced computing resources, for AI/ML workloads.
Yotta has dedicated programmes like Rudra and Shambho to accelerate innovation. Rudra extends Shakti Cloud credits – up to $$50,000 – to Nvidia Inception and Connect programme members, offering access to Nvidia Hopper GPUs and a comprehensive suite of AI tools, including Kubernetes clusters, serverless inferencing, and more. Shambho, launched in partnership with Nasscom and the Telangana AI Mission, provides startups identified through the GenAI Foundry access to up to $200,000 in credits on Shakti Cloud. This initiative is intended to benefit over 3,600 deeptech startups across India.
Additionally, Yotta’s AI Lab as a Service, powered by Nvidia GPUs, provides a powerful virtual lab environment for educational institutions, innovation and emerging hubs — enabling collaborative AI research, model training, and experimentation. Gupta said with AI adoption accelerating, demand for data centre infrastructure will only grow. “We are actively investing in AI-optimised hyperscale data centres across the country – including facilities like Yotta NM1 in Navi Mumbai, D1 in Greater Noida, G1 in GIFT City and H1 in Hyderabad – designed to handle power-dense workloads, optimised for GPU-heavy operations,” he added.
What makes Yotta a significant player is that it will provide over 50% of the advanced GPU compute capacity, soon to be accessible through the India AI Mission portal. “Beyond infrastructure, the company offers AI Labs for students, AI Workspaces, GPU-as-a-service (GPUaaS), and API endpoints for advanced AI models, enabling organisations to develop and deploy customised AI solutions with speed and efficiency.
According to Gupta, there is an urgent need for AI models that can understand, interpret, and generate content in Indian languages – not just in English. “At Yotta, we fully support this national vision by providing the compute infrastructure needed to train and deploy these models sustainably and at scale,” he added.