Jensen Huang, co-founder and CEO of NVIDIA Corp., was held at a press conference in Taipei, Taiwan on Tuesday, June 4, 2024.
Annabelle Chih | Bloomberg | Getty Images
Nvidia CEO Jensen Huang said that the next generation of AI will be 100 times higher than older models because of the new reasoning method that thinks “how to best answer” the question.
Huang told CNBC’s Jon Fortt in an interview with chipmaker Jon Fortt Fourth Quarter Earnings Report.
He cited the models including DeepSeek’s R1, OpenAI’s GPT-4, and Xai’s Grok 3 as models, as models that use the inference process.
The results of the NVIDIA report were that analysts had the highest estimates, with revenues rising 78% a year ago to $39.33 billion. Data center revenues, including NVIDIA’s market-leading graphics processing units or GPUs for AI workloads, surged 93% to $35.6 billion, and now account for more than 90% of total revenue.
The company’s stock is Lost 17% of its value On January 27, it had its worst decline since 2020. The decline is due to concerns raised by China’s AI lab DeepSeek that companies may get higher performance in AI in terms of lower infrastructure costs.
Huang delayed the idea in an interview Wednesday, saying DeepSeek’s popular reasoning model will require more bargaining chips.
“DeepSeek is great,” Huang said. “It’s great because it opens a world-class inference model.”
As the Biden administration ended, NVIDIA was restricted from conducting operations in China due to export controls.
Huang said the company’s revenue percentage in China fell by about half due to export restrictions, adding that there were other competitive pressures nationwide, including Huawei.
Huang said developers might search for ways to revolve around output controls, whether it is for supercomputers, personal computers, mobile phones or gaming consoles.
“Eventually, the software found a way,” he said. “You end up making the software workable on any system you want to locate and create great software.”
Huang said NVIDIA’s GB200 sold in the United States can produce AI content 60 times faster than the chip versions sold to China under export controls.
NVIDIA spends billions of dollars in infrastructure from the world’s largest tech companies every year to earn a lot of revenue from its revenue. The company has been the biggest beneficiary of the AI boom, with revenue more than doubled for five consecutive quarters in mid-2024 before growth slowed slightly.
