Nvidia accelerates in the global AI infrastructure race.

Nvidia forecasts second-quarter revenue of approximately $91 billion, exceeding Wall Street's expectations, indicating that the global wave of artificial intelligence (AI) investment continues to surge despite increasing competitive pressure in the semiconductor market.

Jensen Huang, CEO of Nvidia, speaks at the Milken Institute Global Conference 2026. Photo: Reuters

According to the company's forecast, second-quarter revenue will be significantly higher than the $86.84 billion previously estimated by LSEG analysts. Nvidia also announced an $80 billion share buyback program and an increase in its quarterly cash dividend.

Jensen Huang, CEO of Nvidia, said that revenue from the company's flagship AI chip lines could surpass $1 trillion between 2025 and 2027.

Nevertheless, investors remain cautious amid increasing competitive pressure in the AI chip sector. Nvidia shares fell about 1.6% in after-hours trading following the release of its financial report.

Nvidia's business performance is now seen as a key indicator of the health of the global AI market, as its chips are being used in the majority of large data centers to power advanced AI models.

According to Reuters, spending on AI infrastructure by US tech giants such as Alphabet, Amazon, and Microsoft could exceed $700 billion this year, a sharp increase from around $400 billion in 2025.

Jensen Huang believes Nvidia can still maintain a growth rate faster than the industry's overall expansion of AI spending. According to him, the new customer base of AI-focused cloud service providers is growing faster than traditional technology corporations.

However, Nvidia is also facing increasing competitive pressure as many major customers are simultaneously developing their own chips to reduce their dependence on Nvidia and optimize operating costs.

Analysts believe that the competition in the AI chip market is gradually shifting from model training to "AI inference"—a technology that allows AI to process and respond directly to user requests. This is considered a significantly larger market in the long term.

Besides major technology corporations, Nvidia also has to compete with many semiconductor rivals such as Intel and Advanced Micro Devices (AMD), which are also aggressively pursuing ambitions in the inference AI chip market.

To solidify its leading position, Nvidia introduced the Vera central processing unit in March, along with an AI system developed based on technology from Groq – a startup specializing in inference AI chips.

Jensen Huang said Vera could help Nvidia access an additional market worth approximately $200 billion. The company expects revenue from this chip line to reach around $20 billion in the current fiscal year.

According to Nvidia, this revenue figure was not included in the previous forecast of $1 trillion in revenue from the Blackwell and Rubin AI chip lines.

In the first quarter, Nvidia recorded revenue of $81.62 billion, exceeding analysts' average forecast of $78.86 billion. Data center revenue reached $75.2 billion, higher than the expected $72.8 billion.

The latest business results show that Nvidia remains at the center of the global AI investment wave, even though competitive pressure in the AI chip market is predicted to intensify in the near future.

Source: https://hanoimoi.vn/nvidia-tang-toc-trong-cuoc-dua-ha-tang-ai-toan-cau-751011.html