Nvidia Continues AI Dominance with 427% Growth in Data Center Business

“Customers are putting immense pressure on us to deliver systems quickly,” said Huang. “All the major cloud service providers are consuming every GPU available.”

One standout performer has been Nvidia’s H100 GPUs, with companies like CoreWeave quoting $4.25 per hour for rental. These GPUs are essential for training large language models such as OpenAI’s GPT, demonstrating the critical role Nvidia hardware plays in AI development.

Colette Kress, Nvidia’s Chief Financial Officer, highlighted the strong return on investment (ROI) for their clients. She stated that for every dollar spent on Nvidia hardware, cloud providers could expect to generate five dollars in revenue over four years. Newer hardware, like the HDX H200, promises even greater returns, potentially generating seven dollars for every dollar invested over the same period.

Huang also noted Meta’s significant investment in Nvidia chips. Despite not being a cloud provider, Meta plans to purchase 350,000 Nvidia chips to support its advertising business and potentially incorporate AI chatbots into its apps. This investment underscores the essential nature of AI infrastructure, which Huang referred to as “AI factories.”

Nvidia also announced an accelerated timeline for its next-generation GPU, Blackwell, set to be available in data centers by the fiscal fourth quarter. Early customers for Blackwell include tech giants such as Amazon, Google, Meta, Microsoft, OpenAI, Oracle, Tesla, and Elon Musk’s xAI.

Following the earnings report, Nvidia shares surged by 6% in extended trading, surpassing $1,000 for the first time. The company also announced a 10-for-1 stock split, reflecting its remarkable 25-fold increase in share price over the past five years.