Everyone thinks NVIDIA’s main product is graphics cards. But in reality, its greatest advantage isn’t hardware—it’s the CUDA software platform. In 2006, Jensen Huang bet on parallel computing, which later became the foundation for the development of artificial intelligence. In 2008, the company’s shares plunged by more than 80%, but NVIDIA continued investing in technology. In 2012, researchers trained the AlexNet neural network on NVIDIA graphics cards and made a breakthrough in image recognition. For decades, Huang built the infrastructure that makes it hard to imagine today’s AI industry without it.

Everyone thinks NVIDIA’s main product is graphics cards. But in reality, its greatest advantage isn’t hardware—it’s the CUDA software platform.

In 2006, Jensen Huang bet on parallel computing, which later became the foundation for the development of artificial intelligence. In 2008, the company’s shares plunged by more than 80%, but NVIDIA continued investing in technology.

In 2012, researchers trained the AlexNet neural network on NVIDIA graphics cards and made a breakthrough in image recognition.

For decades, Huang built the infrastructure that makes it hard to imagine today’s AI industry without it.