Copyright © BusinessAMBE 2023

Key takeaways

  • Jensen Huang, CEO of Nvidia, believes that significant capital expenditure on AI infrastructure is sustainable.
  • Leading hyperscalers like Meta, Amazon, Google and Microsoft are driving record spending on AI infrastructure.
  • Nvidia's $10 billion (around €8.5 billion) investment in Anthropic illustrates its commitment to supporting profitable AI innovation by providing the necessary computing power.

Jensen Huang, CEO of Nvidia, reassured investors that the tech industry's huge capital expenditure on AI infrastructure is not only justified, but also sustainable.

Sustainable AI investments

In CNBC's 'Halftime Report' programme, Huang attributed this confidence to the expected increase in cash flows for companies investing heavily in AI. His comments followed recent earnings reports from major Nvidia customers such as Meta, Amazon, Google and Microsoft, which all showed significant increases in spending on AI infrastructure.

Those hyperscalers are expected to spend as much as $660 billion (about €558.5 billion) in capital expenditure this year, a significant portion of which will be spent on buying Nvidia chips. Although Wall Street's reaction to the increase in spending was mixed, Huang remained optimistic. In after-hours trading, Nvidia shares slightly fell to $185.19 at 1.10pm local time.

"Biggest infrastructure expansion ever"

He described the current trend as "the biggest infrastructure expansion in human history", driven by the insatiable demand for computing power that is crucial for AI companies and hyperscalers to generate revenue. Huang gave specific examples of how Nvidia customers are using AI. Meta is switching its recommendation system from CPU-based to generative AI and agents, while Amazon Web Services will use Nvidia chips and AI to improve product recommendations.

Microsoft plans to use Nvidia-powered AI to improve its enterprise software. Huang also praised the performance of leading AI labs OpenAI and Anthropic, both of which use Nvidia chips via cloud providers. Nvidia's $10 billion (about €8.5 billion) investment in Anthropic last year illustrates its commitment to supporting AI innovation.

Profitability as a key indicator

Huang stressed that both Anthropic and OpenAI are highly profitable, illustrating the direct correlation between increased computing power and revenue growth. He noted that all Nvidia GPUs, even six-year-old models like the A100, are currently being leased, indicating continued demand for AI computing power.

Huang concluded that as long as AI remains profitable and companies continue to generate returns on their investments, demand for AI infrastructure will only increase.

Follow Business AM also on Google News

Want access to all articles, enjoy our promo temporarily and subscribe here!

© The Content Exchange, source News