Jensen Huang Nvidia Morgan Stanley TMT Conference

NVIDIA’s transformation from a graphics-focused company to the backbone of global AI infrastructure centers on a full-stack architectural approach. By controlling the entire stack—from chip design and system architecture to libraries and software—NVIDIA maintains a competitive edge in tokens-per-watt efficiency, which is critical as compute capacity directly correlates to corporate revenue and national GDP. The industry is shifting from simple generative models to agentic AI, where systems autonomously reason, use tools, and perform complex tasks. This transition necessitates massive capital investment in "AI factories" that produce tokens as a primary commodity. As software companies evolve into agentic platforms, the demand for compute will continue to scale, positioning NVIDIA at the forefront of physical AI and digital biology, where physical awareness and autonomous capabilities define the next decade of technological advancement.