As it turned out, the introduction of the CUDA software environment in 2006 was an extremely risky business move. The goal of the initiative was to transform the company from a specialized manufacturer of graphics chips into a comprehensive supplier of computing platforms. However, the implementation of the idea was accompanied by huge costs. During that period, the main audience was gamers, who had no demand for advanced programming functions.
The forced implementation of CUDA technology in all consumer devices of the GeForce series led to a 50% increase in production costs and a sharp drop in gross margin to 30%. The market value of the entire company then plummeted to $1.5 billion. The decision created a serious threat to further existence.
Consistent development of the ecosystem over 10 years ultimately paid off. It was the early user base in the PC market that became the company's main advantage in the field of AI. The head of NVIDIA admitted that it was the line of products for gamers that financed and spurred the development of the platform, which today serves the most advanced workstations and servers in the world.