(1) Accumulation stage: In 1993, Huang Renxun co-founded NVIDIA with two young engineers from Sun Microsystem, and devoted himself to the research and development of graphics chips in the early stage. But there were more than 20 graphics chip companies in the market at the time, and three years later that number had soared to 70, with Nvidia not prominent among them. It wasn't until 1997 that Nvidia decided to abandon some of its existing patents in favor of full support for Direct X (the Microsoft standard) that the company launched RIVA 128 in 1997, its first truly successful product.
(2) Rising stage: In 1999, the company launched GeForce 256 and defined the GPU chip, which is the world's first fully functional chip that can truly replace the CPU rendering graphics, and defined the concept of GPU for the first time, thus Nvidia embarked on the road to reshape the graphics card industry. In 2000, the company acquired the graphics card pioneer 3Dfx, once again consolidating its position in the graphics card industry and ATi Corporation formed a duopoly pattern.
(3) Control stage: In 2006, Nvidia innovatively launched CUDA architecture. CUDA, Compute Unified Device Architecture, is a parallel computing platform and programming model based on Nvidia's own GPU. CUDA has two huge impacts. In the GPU industry, CUDA enables Gpus that only do 3D rendering to achieve general computing functions, and the application fields of Gpus can be extended from games (graphics rendering) to high-performance computing, autonomous driving and other fields. As for Nvidia itself, it vigorously promoted CUDA in the early days, and carried out programming language extensions to CUDA, such as CUDA C\/C,CUDA Fortran language, etc., so that developers can easily program GPU. At present, CUDA is one of the two most mainstream GPU programming libraries, laying the foundation for the formation of Nvidia's GPU ecosystem. While Nvidia vigorously promotes the unified platform CUDA and continuously iterates the GPU architecture, its largest competitor ATi was dragged down by its CPU business after being acquired by AMD, and its development was limited, and Nvidia's competitive position in the GPU field was further consolidated at this stage.
(4) Take-off stage: Betting on AI, data center business opens the second growth curve. In 2012, Alex Krizhevsky used GPU for deep learning and won the champion in ImageNet competition after several days of training. He improved the accuracy rate of deep convolutional neural network AlexNet by 10.8%, which shocked the academic circle. This opens the door for GPU applications in deep learning, and it is no surprise that it uses the NVIDIA GTX 580 GPU chip and CUDA computing model. Since then, Nvidia GPU and CUDA models have become the preferred chips for deep learning (especially training links), and Nvidia has also launched a large number of chips and supporting products dedicated to AI, transforming from a graphics card hardware company to an artificial intelligence company. Driven by artificial intelligence, the compound growth rate of the company's data center business from 2014 to 2022 fiscal year reached 64.39%, and its proportion of revenue increased from less than 5% in 2014 to 40% in 2022 fiscal year.
Looking at the history of Nvidia, even if the company has different development priorities at different times, it has always been its excellent innovation ability, strong chip design ability and stable decision execution ability. In fact, Nvidia launches a generation of chip architecture every two years on average, launches a new product every six months, and has persisted for many years, from the Fermi architecture in 2009 to the current Hopper architecture, the company's product performance has steadily improved, and has always led the development of GPU chip technology.