What is Nvidia?

What is Nvidia?

Nvidia is the world's most valuable company — an American semiconductor and AI-computing firm whose data-center GPUs, networking, and CUDA software stack power the overwhelming majority of the world's AI training and inference. Founded in 1993 to bring 3D graphics to PCs, it reported $215.9 billion in revenue for fiscal 2026 (up 65% year over year) and became the first company ever to reach a $5 trillion market capitalization, a level it still held near $5 trillion as of mid-2026.

Nvidia started by selling graphics cards for gaming and pivoted into accelerated computing when it launched CUDA in 2006, turning GPUs into general-purpose parallel processors. When the deep-learning boom arrived, that bet made Nvidia the default platform for AI: its chips, the NVLink and Spectrum-X networking it gained partly through the 2020 Mellanox acquisition, and the CUDA software libraries together form a hard-to-displace full stack.

The business is now overwhelmingly a data-center company. In fiscal 2026 (ended January 2026), Data Center revenue reached $193.7 billion — roughly 90% of total revenue — as cloud providers AWS, Microsoft Azure, Google Cloud, Oracle, and CoreWeave, plus enterprises like SAP, Eli Lilly, and Hyundai, raced to deploy Nvidia's Blackwell systems. Gaming (GeForce) contributed about $16 billion. Demand kept accelerating into the new fiscal year: in Q1 FY2027 (reported May 2026) Nvidia posted record quarterly revenue of $81.6 billion, with Data Center alone at $75.2 billion (up 92% year over year, ~87% of revenue) and networking at a record $14.8 billion, prompting Jensen Huang to say Blackwell sales were 'off the charts' and cloud GPUs were sold out.

Nvidia holds an estimated 80%+ share of the data-center AI accelerator market and ran at roughly 71% gross margin and $120.1 billion of net income in fiscal 2026, on a workforce of about 42,000. Its biggest competitive risk is not a single rival but the rise of hyperscaler custom silicon (Google TPUs, AWS Trainium, Broadcom-designed ASICs) and AMD's data-center GPUs, alongside U.S. export restrictions that limit China sales.

Sources:NVIDIA — Q4 & FY2026 resultsNVIDIA — Q1 FY2027 results (May 2026)companiesmarketcap — NVDA market cap

What does Nvidia offer?

A full-stack AI and accelerated-computing portfolio: data-center GPUs and systems, networking, the CUDA software platform, plus gaming, professional visualization, automotive, and robotics.

Sources:NVIDIA — Data Center productsNVIDIA — homepage product overview

How does Nvidia make money?

Nvidia makes money primarily by selling data-center GPUs, complete AI systems, and networking gear to cloud providers and enterprises, with software and licensing as a growing attach. Hardware is sold per-unit (not per-seat), and the CUDA software moat keeps customers locked to Nvidia silicon.

The vast majority of revenue is hardware sold at high prices: an H100/Hopper-class data-center GPU has typically sold for roughly $25,000–$40,000 per unit, and a rack-scale GB200 NVL72 Blackwell system can run into the low millions of dollars. Customers usually buy through OEMs and cloud providers rather than directly, and a handful of large hyperscalers account for a substantial share of Data Center revenue. On top of the hardware, NVIDIA AI Enterprise software is licensed on a per-GPU subscription — about $4,500 per GPU per year, or roughly $1 per GPU-hour in the cloud — turning the installed base into a recurring software stream.

The unit economics are extraordinary. Fiscal 2026 GAAP gross margin was about 71% and net income was roughly $120.1 billion on $215.9 billion of revenue. Because CUDA and the surrounding libraries are tuned to Nvidia hardware, switching costs are high — the software is the moat that protects the hardware pricing, and it is why even cheaper rival chips struggle to displace Nvidia.

Growth is driven almost entirely by AI-infrastructure buildout: Data Center revenue grew 68% in FY2026 to $193.7 billion and then accelerated to $75.2 billion in a single quarter (Q1 FY2027, up 92% year over year), fueled by the Blackwell ramp across hyperscalers, sovereign-AI projects, and enterprises building 'AI factories.' Consumer gaming (GeForce) and automotive remain meaningful but small relative to the data-center engine, and in May 2026 Nvidia raised its dividend 25-fold and authorized fresh buybacks — a signal of how much cash the model now throws off.

Sources:NVIDIA — Q4 & FY2026 resultsNVIDIA — Q1 FY2027 results (May 2026)NVIDIA — AI Enterprise pricing

Who leads Nvidia?

Co-founder Jensen Huang has been CEO since 1993. His long-tenured executive team includes CFO Colette Kress and worldwide-sales chief Jay Puri, alongside co-founder Chris Malachowsky.

Sources:NVIDIA — corporate leadership / managementThe Org — Nvidia leadership team

How do you contact Nvidia's leadership?

Nvidia's verified email pattern is first-initial + last name at nvidia.com (e.g. jhuang@nvidia.com), the format used by roughly 94% of employees. The addresses below for named executives follow that verified pattern but are not individually published; for official reach use press@nvidia.com (media) or industry-analyst-relations@nvidia.com (analysts).

• Jensen HuangFounder, President & CEOjhuang@nvidia.com
• Colette KressEVP & Chief Financial Officerckress@nvidia.com
• Jay PuriEVP, Worldwide Field Operationsjpuri@nvidia.com
• Press / Corporate CommunicationsOfficial media contactpress@nvidia.com

Sources:RocketReach — Nvidia email formatNVIDIA Newsroom — press contacts

How much funding has Nvidia raised?

Nvidia raised only about $20 million of venture capital before going public in January 1999. It has been a public company on NASDAQ (NVDA) ever since and now carries a market capitalization of roughly $5 trillion (mid-2026) — having become, in late 2025, the first company in history to cross $5 trillion.

Nvidia's private funding was tiny by today's standards. At founding in 1993, Sequoia Capital and Sutter Hill Ventures invested $2 million at roughly a $6 million post-money valuation, after LSI Logic CEO Wilf Corrigan introduced Jensen Huang to Sequoia's Don Valentine. The two firms later added roughly $18 million more, bringing total venture funding to about $20 million before the company ever turned to public markets.

Nvidia IPO'd on January 22, 1999 at $12 per share, valuing the company at roughly $600 million — about a 100x step-up on the original 1993 valuation. That IPO is the last time Nvidia raised primary equity at scale; since then it has funded itself entirely from operating cash flow, and in May 2026 it even raised its dividend 25-fold and authorized new buybacks.

As a public company, Nvidia's 'valuation' is its market cap, which has compounded astonishingly: from ~$560 million at IPO to crossing $1 trillion in 2023, $2 trillion in early 2024, $3 trillion in mid-2024, $4 trillion in 2025, and roughly $5 trillion by late 2025–mid-2026 on the back of the AI boom. With ~$120 billion of net income in FY2026 and tens of billions in cash, the company has no need to raise external capital.

Sources:Sequoia Capital — Nvidiacompaniesmarketcap — NVDA market cap history

How did Nvidia get here?

From a Denny's-table startup in 1993 to the GPU, the CUDA platform, the AI inflection, and a ~$5 trillion market cap.

Sources:NVIDIA — corporate timelineNVIDIA — Q1 FY2027 results (May 2026)Wikipedia — Nvidia

Who are Nvidia's competitors?

Nvidia's rivals span merchant chipmakers (AMD, Intel), custom-silicon enablers (Broadcom, Marvell), and the hyperscalers building their own AI chips (Google TPU, AWS Trainium).

Sources:CNBC — custom AI chips vs NvidiaMotley Fool — biggest threat to Nvidia