Nvidia Displays AI Chips as it Dismisses DeepSeek

Nvidia CEO Jensen Huang is expected to unveil cutting-edge chips for artificial intelligence (AI) and quantum computing on Tuesday, dismissing concerns about China's DeepSeek disrupting the market.

Huang's keynote presentation at Nvidia's annual developers conference is expected to fill the SAP Center in San Jose, California, home of the Sharks NHL hockey team.

Industry analysts anticipate Huang highlighting Nvidia's latest Blackwell line of graphics processing units (GPUs), which includes new updates in the works.

The AI boom propelled Nvidia stock prices to unprecedented heights until a sharp sell-off early this year, triggered by DeepSeek's unexpected success.

Despite a recent rebound from a March low, the stock, one of the most traded on Wall Street, has fallen more than 9% this year.

DeepSeek, based in China, shook up the world of generative artificial intelligence (GenAI) with the release of a low-cost, high-performance model that challenges OpenAI and other big-spending behemoths.

However, several countries have questioned DeepSeek's handling of data, which the company claims is collected on "secure servers located in the People's Republic of China."

Nvidia's high-end GPUs are in high demand from tech giants building data centers to power artificial intelligence, and some believe a low-cost alternative could harm the Silicon Valley chipmaker's business.

Ben Van Roo, co-founder and CEO of Yurts, which specializes in protecting sensitive data while allowing AI models to access it, believes DeepSeek's popularity bodes well for Nvidia.

"DeepSeek drastically accelerated the desire to consume these models," Van Roo told AFP.

"You've opened the world's appetite even more (to generative AI) and independent of the fact that it's Chinese, I think it was a good day for Nvidia."

Blackwell is booming

Nvidia has increased production of its top-of-the-line Blackwell processors for AI, generating billions in sales in its first quarter on the market.

"AI is advancing at light speed" and is laying the groundwork "for the next wave of AI to revolutionize the largest industries," Huang told financial analysts recently.

Huang believes Nvidia chips and software platforms will continue to power or train AI for robots, cars, and digital "agents," which refers to AI that can make decisions instead of humans.

The CEO is also likely to discuss a transition to quantum computing.

Following several failed predictions, quantum computing is accelerating rapidly, with actual use cases and scientific breakthroughs expected in years rather than decades.

US tech behemoths, startups, banks, and pharmaceutical companies are all investing in this revolutionary technology.

For instance, Nvidia GPUs' ability to manage several computing tasks concurrently makes them ideal for quantum computing.

Washington has restricted the technology's exports, and China and the United States are leading the way in quantum development.

Demand for Nvidia's chips to power artificial intelligence in data centers drove the company's record $130.5 billion in revenue last year.

Nvidia expects revenue of $43 billion in the current fiscal quarter, exceeding analyst expectations.