Nvidia on Wednesday said its profit soared to $12.3 billion in the recently ended quarter on record high revenue driven by demand for its chips to power artificial intelligence.

The Silicon Valley chip titan reported profit of $12.3 billion on record revenue of $22.1 billion in the quarter than ended late January and record high revenue of $60.9 billion for the fiscal year.

Profit in the same quarter a year ago was $1.4 billion, while revenue in the same three-month period a year earlier was $6 billion, the company said.

"Accelerated computing and generative AI have hit the tipping point," Nvidia chief executive Jensen Huang said during an earnings call.

"Demand is surging worldwide across companies, industries and nations."

Nvidia shares shot up more than eight percent to $732.99 in after-market trade that followed the release of the earnings results.

"The AI Revolution starts with Nvidia and in our view the AI party is just getting started," Wedbush analyst Dan Ives said in a note to investors.

Wedbush expects 60 percent or more of businesses to "head down the AI use case path" in the coming decade, spending an estimated trillion dollars on the technology, according to Ives.

Money taken in by the Nvidia unit specializing in data center computing hit a record of $18.4 billion in the quarter, more than quadruple the amount from the same period a year earlier, according to Nvidia.

"Our Data Center platform is powered by increasingly diverse drivers," Huang said.

"Vertical industries — led by auto, financial services and healthcare — are now at a multibillion-dollar level."

The company said it expects total revenue of $24 billion in the current quarter.

AI factories

Huang said Nvidia chips are at the heart of datacenters transforming into "AI generation factories" with data a "raw material" turned into experiences such as prompt-generated videos made using OpenAI's new Sora tool.

In regions outside the United States, "sovereign AI" systems are adding to demand for Nvidia chips, according to chief financial officer Colette Kress.

She described sovereign AI as large language models made using regional data and supported by local research and business "ecosystems."

Sovereign AI infrastructure is being bult in Canada, France, Japan and many other countries, according to Huang.

"The reason for sovereign AI has to do with the fact that the language, the knowledge, the history, the culture of each region are different and it belongs to them," Huang said on the earnings call.

"They would like to use their data to create their own digital intelligence."

China reset 

Demand for Nvidia chips for datacenters was strong everywhere except China, where revenue dropped significantly following ramped up US export controls imposed in October of last year, according to Kress.

Calls to further close the supply chain grew after the world discovered the powers of AI with the launch of ChatGPT, a tool that debuted in November 2022.

Also causing alarm in Washington was news that China-owned Huawei had released a new smartphone that featured a powerful home-grown advanced chip.

When announcing the beefed-up curbs, US officials insisted they were intended to close loopholes and prevent China's development of AI for military use.

China said at the time that it was "strongly dissatisfied" and "firmly opposes" the curbs.

Nvidia paused shipments of restricted products to China, but is exploring selling alternatives that don't require licenses, according to company executives.

"We reconfigured our products in a way that is not software hackable, and that took some time," Huang said.

"So we reset our product offering to China and now we're sampling to customers in China."

US export control regulations aimed at China and other markets including Vietnam and parts of the Middle East were expected to continue to cause sales of Nvidia data center chip to suffer in those markets.