Nvidia a Timeline From Graphics Cards to AI
Nvidia did not pivot into artificial intelligence. This video traces the thirty-three years that came before AI needed it: a graphics-card company founded on 5 April 1993 by Jensen Huang, Chris Malachowsky and Curtis Priem; the 1999 GeForce 256, marketed as 'the world's first GPU'; and CUDA, launched in November 2006 to open GPUs to general-purpose scientific computing, years before anyone outside the company thought about artificial intelligence.
What followed: three new chip generations aimed at the AI data-centre market - Hopper (2022), Blackwell (2024), Rubin (2026) - and a market-capitalisation climb with no precedent: $1 trillion (30 May 2023, first US chipmaker), $2 trillion (23 February 2024), $3 trillion (5 June 2024, passing Apple), $4 trillion (9 July 2025, first company ever), $5 trillion (29-30 October 2025, first company ever), and $5.5 trillion (13 May 2026, a new record). Fiscal 2026 revenue: $215.9 billion, $193.7 billion of it from Data Center. 42,000 employees as of January 2026.
Chip-architecture technical detail - node names, die sizes, precision-format performance figures - belongs to this channel's companion video and is not repeated here. This video's subject is the company's history, not one generation's engineering.
Educational documentary. Not financial or investment advice.
In these topics
Tags
Chapters
- One idea, kept for thirty years
- A restaurant booth, 1993
- Coining a name, 1999
- A door nobody was aiming at, 2006
- The competition that changed the question, 2012
- Three names, one direction: 2022 to 2026
- What the stock market decided, 2023 to 2026
- The scale behind the number
- What thirty years of one idea adds up to
Sources and credits
Primary sources
- Nvidia, 'Our History: Innovations Over the Years', nvidia.com/en-us/about-nvidia/corporate-timeline - founding date (5 April 1993), founders, 1999 GPU claim, 2006 CUDA launch, 2012 AlexNet framing.
- Nvidia, 'NVIDIA Announces Financial Results for Fourth Quarter and Fiscal 2026', nvidianews.nvidia.com, published 25 February 2026 - full-year revenue $215.9 billion (up 65%), Data Center segment revenue $193.7 billion (up 68%).
- Nvidia developer blog, 'NVIDIA Hopper Architecture In-Depth', 22 March 2022, and 'CUDA Refresher: Reviewing the Origins of GPU Computing' - Hopper announcement date, CUDA background.
- Krizhevsky, Sutskever, Hinton, 'ImageNet Classification with Deep Convolutional Neural Networks', NeurIPS 2012 proceedings - AlexNet, authors, two Nvidia GTX 580 GPUs used for training.
- TweakTown, PC Gamer, VideoCardz, 25th-anniversary coverage of the GeForce 256, cross-checked - release date 11 October 1999.
- CNBC, 30 May 2023 ($1 trillion); CNN/CNBC/Forbes/Axios, 23 February 2024 ($2 trillion); CNBC/Al Jazeera/AIBusiness, 5-6 June 2024 ($3 trillion); CNBC, 9 July 2025 ($4 trillion); CNBC, 30 October 2025 ($5 trillion); Forbes, 13 May 2026 ($5.5 trillion) - market-capitalisation milestones, each cross-checked across at least two outlets.
- CNBC, 30 October 2025, and NBC New York (Denny's founding story, cross-checked across outlets, reported as Huang's own account, not an independently documented fact).
- Nvidia, Form 10-K for fiscal 2026 (period ended 25 January 2026), sec.gov - approximately 42,000 employees.
Not regulated financial advice.