According to Data Center Dynamics, NVIDIA CEO Jensen Huang said at the AI Summit in Mumbai that the American company has entered into a series of partnerships with Indian firms to deploy its chips and AI technologies, expanding its presence in a key market. Some of the firms have been contracted to supply tens of thousands of H100 accelerators, notably Tata Communications and Yotta Data Services.
Tata Communications is upgrading its cloud AI infrastructure in India. The company will begin the first phase of large-scale deployment of NVIDIA Hopper later this year, and will add NVIDIA Blackwell accelerators in the 2025 phase. According to Tata Communications, its platform will be one of the largest AI supercomputers in India. Yotta Data Services has already introduced six new AI services for its Shakti Cloud platform, including those based on NVIDIA NIM. These include AI Lab, AI Workspace, Serverless AI Inferencing, GPUaaS, etc.
Image source: NVIDIA
NVIDIA is also working with the country’s largest conglomerate, Reliance Industries, to create a cloud-based AI infrastructure to process data, train employees, and create its own large language models that support common languages in the country. As part of the partnership, Reliance will deploy GB200 super accelerators. Tech Mahindra intends to use NVIDIA chips and software to develop a Hindi AI model called Indus 2.0. Work on this project will be carried out at the Center of Excellence based at Tech Mahindra’s laboratories in Pune and Hyderabad.
Last week, Jensen Huang also attended the launch ceremony of sovereign AI supercomputer Gefion. According to The Register, in his speech Huang noted that the EU should accelerate progress in the field of AI. “Every country is awakening to the understanding that data is a national resource,” said NVIDIA CEO. European countries need to invest more in AI if they want to close the gap with the US and China, the head of NVIDIA said, implying that the best way to do this would be his company’s accelerators.