In a major push to tap into India's rapidly growing AI market, Nvidia announced a series of new partnerships with prominent Indian companies and launched a lightweight artificial intelligence (AI) model tailored for the widely-used Hindi language. The announcements were made during an AI summit hosted by Nvidia in Mumbai, featuring CEO Jensen Huang and Reliance Industries Chairman Mukesh Ambani.

During the event, Huang emphasized the significant expansion of Nvidia's AI footprint in India, highlighting plans to increase computing infrastructure in the country nearly 20-fold by the end of the year. “Nvidia is AI in India,” Huang stated. “In just one year's time, we will have nearly 20 times more compute here in India than just a little over a year ago.”

As part of its strategy, Nvidia introduced a new AI language model, Nemotron-4-Mini-Hindi-4B. With 4 billion parameters, the model is designed to help Indian firms develop AI applications tailored for local languages. Nvidia said the model was trained using a combination of real-world Hindi data, synthetic Hindi data, and an equal portion of English data. This positions the model to cater to India's diverse language landscape, where only a small percentage of the 1.4 billion population speaks English, and 22 languages are constitutionally recognized.

Indian tech giant Tech Mahindra became the first to adopt Nvidia’s latest model to create Indus 2.0, an AI model centered around Hindi and its various dialects. In addition to this, Nvidia has strengthened collaborations with major Indian IT companies such as Infosys, TCS, and Wipro to train half a million developers in designing and deploying AI applications using Nvidia's software.

Reliance and Ola Electric have also joined forces with Nvidia to utilize its "Omniverse" simulation technology, which allows companies to test factory designs in a virtual environment, enhancing efficiency and reducing costs.

The AI summit in Mumbai saw a turnout of several thousand attendees, highlighting the rising interest in AI in India. The event experienced a slight delay due to the large crowd eager to see Huang, a prominent figure in the AI sector.

Unlike the large language models like OpenAI’s GPT-4, Nvidia's new small language models are focused on specific datasets, making them more affordable and practical for companies with limited resources. This makes them particularly appealing in markets like India, where there is a growing demand for localized AI solutions.

Nvidia's investment in India comes as global chip manufacturers look to strengthen the country's semiconductor industry, aiming to establish India as a significant player in the global tech landscape. Nvidia has maintained a presence in India for nearly two decades, with engineering and design centers in key cities like Bengaluru and Hyderabad.

This latest announcement builds on Nvidia’s previous partnerships, including a 2023 agreement with Reliance to develop AI supercomputers in India and a collaboration with the Tata Group to create AI models for Indian languages.