One of the biggest questions facing Nvidia (NVDA +0.67%) in recent years has been the timing of the company's return to the Chinese market. It's been nearly a year since the Trump administration unceremoniously banned the export of artificial intelligence (AI) chips to customers in the country, only to reverse course months later.
The Chinese government responded swiftly by banning the use of U.S.-made AI chips in government facilities and state-sponsored organizations. The Trump administration ultimately approved the sale of AI-centric graphics processing units (GPUs) to China, contingent on the seller sharing 25% of the proceeds with the U.S. government.
After months of back-and-forth, Nvidia is preparing for a triumphant return to the AI chip market in China.
Image source: The Motley Fool.
A startling revelation
As recently as last month, Nvidia was throwing cold water on the prospects of a return to the world's second-largest market. During the earnings call held on Feb. 25, CFO Colette Kress said, "While small amounts of H200 products for China-based customers were approved by the U.S. government, we have yet to generate any revenue, and we do not know whether any imports will be allowed into China."
That outlook changed this week with a surprise announcement by CEO Jensen Huang at Nvidia's GPU Technology Conference (GTC) on Monday. "We've been licensed for many customers in China," he said. "We've received purchase orders from many customers, and we're in the process of restarting our manufacturing. Our supply chain is getting fired up."
While estimates vary, Nvidia generated revenue of $12 billion to $15 billion from China in 2024, the last full year of sales in the country. Future sales could be much higher, as Huang previously estimated the size of the market at $50 billion.
NASDAQ: NVDA
Key Data Points
There's more. Nvidia is developing a version of its Groq AI chips for sale to Chinese customers, according to a report that first appeared in Reuters. Late last year, Nvidia acquired inference chip assets from the start-up for $20 billion. Inference is the application of AI models to a specific task after they have been trained. Nvidia is developing a version that can be modified to work with other systems, according to the report.
At GTC this week, the company unveiled the highly anticipated Nvidia Groq 3 AI chip, designed to accelerate inferencing. The company's GPUs can be used for both AI training and inference, but rivals have made inroads into Nvidia's dominance by developing solutions that are more efficient for inferencing tasks.
Analysts' consensus estimates already called for Nvidia to generate $368 billion in revenue over the next year, an increase of 70%. Any sales to China would push that forecast higher.
And Nvidia's valuation is compelling at just 22 times forward sales, particularly for a company with such a long runway of growth ahead.
