Nvidia Beats Estimates With $81.62 Billion in Q1 Revenue; Shares Up 0.2% After Hours

Nvidia, the dominant supplier of artificial intelligence chips, reported quarterly results that topped Wall Street estimates and issued a strong revenue forecast, underscoring resilient demand for AI chips.

Nvidia said in an earnings report released after the New York market close on May 20 that fiscal first-quarter revenue for February through April was $81.62 billion, while earnings per share came in at $1.87.

That exceeded market expectations of $79.18 billion in revenue and $1.77 in earnings per share.

The company’s outlook for the current quarter also topped Wall Street forecasts.

Nvidia projected fiscal second-quarter revenue of $89.1 billion to $92.8 billion for the quarter ending in July.

That was above the market estimate of $87.3 billion.

Nvidia shares ended the regular session up 1.3%. The stock briefly fell more than 2% in after-hours trading immediately after the earnings release as investors judged lofty expectations had already been priced in. As of 4:30 p.m., the shares were up 0.2%.

Revenue from Nvidia’s core data center business totaled $75.2 billion.

That topped the market consensus of $73.47 billion and nearly doubled from $39.11 billion a year earlier, extending the unit’s rapid growth.

Buoyed by the strong results, Nvidia also said it would raise its quarterly dividend to $0.25 a share.

Nvidia’s growth may continue in the near term as spending on AI infrastructure expands. Over the longer term, however, big tech companies’ push to develop in-house chips could threaten its dominant position.

Cerebras Systems, which recently completed an initial public offering, is seeking to enter the market with AI processors built around large silicon wafers. AMD is preparing to launch a large-scale AI server system later this year.

Amazon said in its latest earnings report that its in-house chip business had surpassed a $20 billion annualized revenue run rate and was posting triple-digit growth.

Amazon Web Services has signed a deal with OpenAI to provide 2 gigawatts of infrastructure for its Trainium AI chips. It has also agreed with Anthropic on a deal for as much as 5 gigawatts of Trainium chip usage.

Google is also accelerating the expansion of its own AI semiconductor ecosystem. At its annual Google I/O developer conference this week, the company unveiled the TPU 8i chip for AI inference and the TPU 8t for model training.

Park Sang-kyung, Hankyung.com reporter highseoul@hankyung.com