Nvidia CEO Jensen Huang has noted that The United States is no longer clearly ahead of China. in the race for artificial intelligence. His statements, made in an interview with CNBC and reported by various media outlets, point to a much tighter picture than previously assumed.
In his analysis, the manager emphasizes that rapid adoption and abundant energy are driving China's advance, along with open-source models, while the US retains a clear advantage in cutting-edge chip design. At the same time, global demand for computing to train AI systems has skyrocketed in recent months, with a notable boost in the new generation of Nvidia, Blackwell.
An increasingly tight leadership
Huang maintains that the gap between the two powers has narrowed rapidly and that, although The US continues to set the pace on certain frontsThe gap is no longer wide. He cites as examples the push for open models in China and the pace of deployment of new applications.
According to the executive, Chinese open systems —with actors such as DeepSeekAlibaba or Baidu—are moving full steam ahead, while American models remain at the forefront of performance. The battle, however, is increasingly being fought at the user level: The advantage could be decided by who deploys and scales applications first. from day to day.
Huang points out that the Demand for AI computing has grown strongly in the last semester, and that the global board changes in step with that capacity.
In this regard, the CEO emphasizes that China is making progress in infrastructure and software while expanding its user and developer base, closing the gap with the North American power.
Energy, hardware and models: who has the advantage
One of the points that weighs the most is energy: according to figures cited by various reports, China could exceed 10.000 TWh of electricity by 2024, more than double that of the U.S. That electrical muscle powers large-scale data centers, key to model training.
In high-end hardware, the US leads chip design with families like Nvidia Blackwell. Even so, Huang points out that China “is not running out of options”: companies like Huawei They are driving the Ascend series and giants like Alibaba and Baidu are already training systems with self-developed chips.
The push of the open source models strengthens the Chinese ecosystem, which accelerates adoption in multiple sectors thanks to less restrictive regulatory frameworks in certain industries. This rapid deployment is, according to Huang, a factor that reduces time and about application results.
In parallel, the market reflects this narrative: investors have bet heavily on Chinese technology companies and, according to data collected in the economic press, Alibaba is said to have rebounded by around 180% and Xiaomi by 125% year-on-year., in a context of expectations about domestic chips and AI.
US Strategy: Openness, Talent, and Where to Compete
Huang warns that close the export of technology can be counterproductiveIf American innovation doesn't diffuse, other regions will fill the gap. He cites an idea shared by AI advisors at the White House: if the US technology stack represents 80% of the global market, the country is winning; if it falls to 20%, it is losing ground.
To sustain this presence, he mentions the role of companies such as Azure, CoreWeave or Anthropic, which would be driving the international deployment of services and platforms built on American technology. The goal: for developers and companies prefer to build on that infrastructure.
Talent is the other pillar. Huang is in favor of maintaining the visa. H‑1B to attract highly qualified profiles, warning that making it more expensive - like the $100.000 visa fee introduced in September by Donald Trump— makes it difficult to attract specialists and may reduce the competitiveness of US-based multinationals.
In strictly industrial terms, Nvidia is experiencing a strong pull of the Blackwell familyThe company, according to the executive, is detecting very high orders for AI training and, in addition, look for startups to invest in with an eye on the next wave of applications. In this context, the relationship with the ecosystem of OpenAI, which some reports place at large-scale agreements.
In terms of adoption, the Asian country is stepping on the accelerator: a plan from the Council of State indicates the goal of 70% of the population will be using AI applications by 2027. It is also noted that China has more than a billion users, concentrating around 50% of AI researchers and retains around 30% of the global technology market, figures that underline its structural weight.
For Huang, the “new industrial revolution” will be decided at the application layer, where models, data and services quickly move from prototypes to products. Whoever deploys first already scales will make a difference in both productivity and technological influence.
The general outline left by the manager is unequivocal: the gap exists but it is shortChina is advancing in energy, data centers, and open source; the United States retains the lead in high-performance chips and global platforms. The outcome will depend on market opening, talent attraction, and the speed with which research is converted into massive AI applications.
