How Nvidia, Huang fit into US-China AI race | Number Theory | Hindustan Times

The first part of this series discussed Nvidia’s rise as one of the most valuable companies in the world and how it matters for the larger AI economic story. The concluding part will bring in the geopolitical in the economic AI story, largely the Sino-US competition.

BloombergChina has won the trade war, can it also win the AI war? 2026 will be the second consecutive year when China will post a trade surplus above a trillion dollars. This kind of mercantile dominance is unprecedented in the world. China today, is not just the cheap factory of the world, like it was a few decades back. It controls some of the most cutting-edge technology and manufacturing processes in the world, so much so that even European and American giants in many sectors are now lagging. Even on AI, as was shown in an HT analysis of a Carnegie China report, China is catching up pretty fast with the US in terms of the skilled workforce in the sector. The world realised this even before the Carnegie China report came, when China released DeepSeek-R1– an open-source reasoning model – in January 2025, bringing it within striking distance of all leading frontier AI developers, all of whom are based in the US. A recently published report by Lowy Institute also shows that Chinese models are catching up fast with American models despite the US having a lead over China in every AI field except human resource.But China still needs advanced chips for its AI push Jensen Huang has been seen as a villain by many US hawks because he has also been a votary of the relaxing restrictions on sale of Nvidia chips to China. The latest controversy around his extra-marital affair and the allegations of Alice Huang being an alleged Chinese spy have only added to the hawkish views around Huang. China, on expected lines, has downplayed the controversy and denied the allegations of any kind of spying. While a lot of commentators argue that China does not really need Nvidia’s chips – the actual evidence is anything but clear – an HT analysis of trade data shows that unlike in most things, China does show a rising import dependence in commodities which are likely GPUs or related products. This is where Nvidia, both in terms of its chip supplies and technical know-how in terms of any alleged corporate espionage via Huang himself being compromised, becomes important.The AI-race as of now is contingent on the US and China solving their respective supply constraints first The easiest way to describe the US and China trying to derail or constrain each other’s progress in the AI race is the US trying to restrict China’s access to cutting edge chips/GPUs and China limiting the supply of rare earth minerals to the US. Both sides perhaps know that the constraints will cease to matter in the long-term but they hope to crack the challenge before they become redundant. “Washington has imposed numerous restrictions on both China and Chinese subsidiaries abroad, which has, in turn, spurred affected companies to innovate around them. In May 2026, Huawei announced a new chip-design approach intended to increase transistor density without shrinking transistors. This could allow it to design chips by 2031 with transistor density equivalent to that of the chips Taiwan’s TSMC aims to make in 2028, but without the advanced lithography machines China is barred from importing. If successful, this could limit the effect of US export controls”, the Lowy Institute report says. The US, on the other hand, launched the Pax Silica alliance in December 2025 which will ease its access to rare earth minerals. Beijing countered the move with World Artificial Intelligence Cooperation Organization (WAICO) earlier this year. All these are independent of Huang’s affair but the disclosure of the latter is certain to add more paranoia in the US AI eco system vis-à-vis fears of a Chinese catch-up in the biggest technological battle of our age.ABOUT THE AUTHORRoshan KishoreRoshan Kishore is the Data and Political Economy Editor at Hindustan Times. He heads the newsroom's data journalism team, which produces Number Theory, a daily data-driven feature for the print edition and the HT app. Number Theory uses data analysis and story-telling based on it to add value to the newsroom’s daily coverage by putting stories in a larger context on a range of issues, including politics, macroeconomy, markets, global affairs and climate. Under his leadership HT’s data journalism work has established itself as a niche product in Indian journalism and pushed the boundaries of marrying academic rigour with news sense and speed. Along with writing and editing data stories, he has also been writing a weekly political economy column called Terms of Trade for HT Premium. A trained economist with an MPhil degree from Jawaharlal Nehru University, Kishore has also been a visiting fellow at the Centre for Advanced Studies of India (CASI) at the University of Pennsylvania. Along with his journalistic work, his writings have also appeared in journals such as the Economic and Political Weekly and working papers for CASI and UNESCAP.Read More

Abhishek Jha is Assistant Editor-Data at Hindustan Times. He uses statistical programming to generate newsworthy insights from large datasets. He is part of the team that produces Number Theory, a daily data story feature of the paper’s print edition. Since March 2024, he has been writing Weather Bee, a weekly column for the Hindustan Times website. He is a chemical engineer by training, who specialises in stories related to weather, climate, and the environment. Jha has been at HT since 2018, where he offers data-driven perspective and analysis on politics, environment, weather, climate, economy and society. His work includes data coverage of elections in India and abroad, including the 2019 and 2024 Lok Sabha elections; the disasters and extreme weather resulting from changing climate, such as floods, droughts, heat waves, cold waves, and dwindling snow cap in the Himalayas; the factors that drive poor air quality in northern India; the changing patterns of land use; the Covid-19 pandemic and its impact on labour market conditions; the changing pattern of consumer spending seen in the new consumer spending surveys; and social norms seen in the surveys such as the National Family Health Survey.Read More

Shreya Khandelwal is a trainee (data) at Hindustan Times and part of the team that produces Number Theory, a daily data story feature of the paper’s print edition. Her primary responsibility on the team is to cover economic stories using publicly-available datasets. She also contributes to reporting on the economy for the newspaper's business section. Before joining the Hindustan Times newsroom in June 2026, she completed her MA in Economics from Azim Premji University and BA (Prog.) in Economics and Mathematics from Jesus and Mary College, University of Delhi. During her master's, she interned at the Centre for Analytical Finance (Indian School of Business), where she provided inputs on the industrial sector for the Telangana Rising 2047 vision project and the Council on Energy, Environment and Water (CEEW) where she worked on industrial pathways in developed and developing countries and foreign trade agreements.Read More