Two GTX 580s in SLI are responsible for the AI we have today — Nvidia's Huang revealed that the invention of deep learning began with two flagship Fermi GPUs in 2012

Nvidia CEO Jensen Huang revealed that the inception of deep learning, pivotal for AI development, traces back to researchers using two GTX 580 GPUs in SLI in 2012. The researchers at the University of Toronto crafted AlexNet, a groundbreaking image recognition architecture with self-learning capabilities, outperforming existing algorithms by over 70%. This development marked Nvidia's entry into AI hardware, leading to subsequent innovations like the original Nvidia DGX and Volta architecture. The use of GTX 580s for deep learning laid the foundation for Nvidia's significant investments in AI technology, shaping its current standing in the industry.