Nvidia Corp. Chief Executive Officer Jensen Huang on his visit to India earlier this month had a one-on-one conversation with Prime Minister Narendra Modi about the AI sector.
As the US increasingly clamps down on exports of high-end chips to China and the world seeks an alternative electronics manufacturing base, India could shape up to be a source of AI talent, site for chip production and market for Nvidia’s products. Nvidia, whose graphics processors are vital to the development of artificial intelligence systems.
While in Delhi, Huang spoke of re-training entire swaths of the country’s workforce and building future AI models using Indian data and talent. Huang also told one executive in Bangalore that he’s a big believer in the country’s engineering talent, particularly graduates from its IITs.
Huang added: “This is going to be one of the largest AI markets in the world.”
Nvidia and India have a shared interest in betting on, and speeding up, the country’s AI ascendancy. Chipmakers cannot sell top-end microprocessors to China, which accounts for a fifth of Nvidia’s sales, amid fears the chips could be used to develop autonomous weaponry or wage cyberwarfare. “India is the only market remaining so it isn’t surprising that Nvidia wants to put multiple eggs in that basket,” said Neil Shah, vice president of research at Counterpoint Technology Market Research.
While Indian engineers are a vital part of the digital workforce, the country is still far away from developing the cutting-edge capabilities needed to manufacture Nvidia’s sophisticated chips. But India is plowing billions in subsidies to set up chip manufacturing infrastructure to lure the likes of Nvidia, Advanced Micro Devices Inc. and Intel Corp.
“India is strategic to Nvidia’s future,” said Nandan Nilekani, chairman of Infosys Ltd. and a designer of vital portions of the country’s massive digital public infrastructure. “The government is aggressively building AI infrastructure and so are large private companies. That’s good news for Huang,” said Nilekani.
India’s largest conglomerate, Reliance announced during Huang’s multi-city tour that its Jio Platforms would build AI computing infrastructure for the country. The AI cloud will use end-to-end Nvidia supercomputing technologies, the company said in a release. Reliance and another large conglomerate, Tata, will also build and operate state-of-the-art AI supercomputing data centres and offer AI infrastructure as a service for use by researchers, corporations and startups, Nvidia said.
India has had reasonable success in enticing giants Apple Inc. and Amazon.com Inc. to shift contract electronics manufacturing from China – to a point where, this month, Apple will sell India-made iPhone 15 devices on launch day. Now it’s turning to semiconductors, armed with some experience in chip design and no history whatsoever in semiconductor foundries. Nearly all cutting-edge chips, including those designed by Nvidia, are made in Taiwan. Over several decades, the country spent billions to reach current levels of manufacturing sophistication.
India wants to catch up, but it faces challenges turning itself into an AI hub. The country currently has no exascale compute capacity – being able to handle a billion billion calculations per second – nor the ready AI talent capable of writing sophisticated software, said Sashikumaar Ganesan, who chairs the computational and data sciences department at Indian Institute of Science. “We have to build not just AI infrastructure but also a high-performance computing workforce,” said Ganesan.
Still, India is a fast-maturing market for high-end technologies. That’s created huge demand for Nvidia’s GPUs, or graphics processing units. As India’s digital economy grows, the government is mandating data security, data privacy and data localization, and this could require over 100,000 GPUs to build AI cloud infrastructure.
Nvidia already has four engineering centres in India, including in Bangalore and in Gurugram, with a total of 4,000 engineers, its second-biggest talent pool after the US.