Nvidia

What Nvidia does

Nvidia designs graphics processing units (GPUs). These chips grew out of PC gaming graphics and are now widely used to train and run artificial intelligence models. The company reports two segments in its annual report for fiscal 2026. It says Compute & Networking covers data center and automotive platforms. Graphics, it says, includes GeForce chips for gaming PCs and RTX chips for workstations. Nvidia outsources manufacturing. The filing says it uses foundries such as TSMC and Samsung to produce its wafers. It says Nvidia had about 42,000 employees in 38 countries at the end of the fiscal year.

Why it matters

Most of Nvidia's revenue comes from data centers. Fiscal 2026 revenue, for the year ended January 25, 2026, was $215.9 billion, up 65%, of which $193.7 billion was data center revenue, the company reported. On September 21, 2026, The Register described Nvidia as the world's most valuable company, with a market value of nearly $5.5 trillion.

TechCrunch, citing CNBC, reported in May 2026 that Nvidia had committed more than $40 billion that year to equity stakes in AI companies, including $30 billion in OpenAI. It quoted a Wedbush Securities analyst who placed the deals within a circular investment pattern, in which Nvidia funds companies that buy its products. Chief executive Jensen Huang rejected that label at an investor conference in September 2026, according to TechCrunch. He said far more money comes back than Nvidia puts in.

Recent developments

In April 2025 the US government told Nvidia it needed a license to sell its H20 chip in China, the annual report says. It says the company recorded a $4.5 billion charge. Licenses followed in August 2025. The filing says US officials expected the government to receive 15% or more of the licensed revenue, though no regulation codified that. In January 2026 the Commerce Department set conditions for exporting Nvidia's H200 chip to China, including a cap at 50% of US shipment volume, The Register reported. Nvidia's August 2026 outlook assumed no data center compute revenue from China.

In December 2025, according to TechCrunch's timeline of 2025, Nvidia:

• struck a non-exclusive licensing deal with chip startup Groq
• hired its founder and other staff
• bought $20 billion of Groq assets

In September 2026 The Register reported, citing the New York Times, that the Justice Department had opened an antitrust investigation into the arrangement. Nvidia defended the deal in a statement quoted by The Register.

At its GTC conference on March 16, 2026, Huang said he saw at least $1 trillion in orders for the Blackwell and Vera Rubin chip generations through 2027, TechCrunch reported. For the quarter ended July 26, 2026, Nvidia reported revenue of $96.2 billion, up 106% from a year earlier. On September 3 Huang announced an agreement to acquire Hugging Face, which hosts AI models and datasets, for about $12.9 billion. On September 28 Nvidia added $150 billion to its share buyback authorization.

What to watch

The Hugging Face purchase is expected to close in the first half of 2027, subject to regulatory approval, according to The Register. It quoted a Forrester analyst saying the platform's value comes from its neutrality. Huawei's rotating chairman Eric Xu claimed in September 2026 that its Ascend chips had surpassed Nvidia's market share in China, based on Huawei's own data, The Register reported. Amazon, Google, Microsoft and OpenAI are among the companies building their own AI chips, TechCrunch reported on August 31, 2026. It said Nvidia was investing $3.5 billion in MediaTek, which will design custom chips that plug into Nvidia-based data centers.