Nvidia founder and CEO Jensen Huang recently spoke at SIGGRAPH in Los Angeles about the company’s decision to embrace AI-powered image processing in 2018, which has redefined its future and helped to redefine an evolving industry. Huang stated that the choice to embrace ray tracing and intelligent upscaling was a “bet the company” moment that required the reinvention of hardware, software, and algorithms. This decision has paid off enormously, and Huang believes that this is only the beginning of an AI-powered near future, which will be powered primarily by Nvidia hardware.
The Rise of Ray Tracing and Intelligent Upscaling
Huang explained that rasterization, the traditional method of rendering a 3D scene, was reaching its limits. Ray tracing and intelligent upscaling, known as RTX and DLSS, respectively, were still in the process of being adopted across the diverse and complex world of consumer GPUs and gaming. However, the architecture that Nvidia had created to enable these technologies was found to be a perfect partner for the growing machine learning development community.
The Future of AI
Huang believes that the future is an LLM at the front of just about everything, and that “human” is the new programming language. He predicts that everything from visual effects to a rapidly digitizing manufacturing market, factory design, and heavy industry will adopt, to some degree, a natural language interface. Huang also believes that entire factories will be software-defined and robotic, and the cars they’ll be building will themselves be robotic. He stated that “it’s robotically designed robots building robots”.
The Need for Computing Resources
The new models not only need to be trained but also run in real-time by millions, perhaps billions, of users on a regular basis. Huang believes that investing millions of dollars in last-generation computing resources, like CPU-focused racks, is foolish when something like the GH200, the newly revealed and datacenter-dedicated AI development hardware, can do the same job for less than a tenth of the cost and power requirements. Huang presented a video showing a LEGO-like assembly of multiple Grace Hopper computing units into a blade, then a rack, then a row of GH200s all connected at such high speeds that they amounted to “the world’s largest single GPU,” comprising one full exaflop of ML-specialty computing power.
Conclusion
Huang outlook for the future of AI is optimistic, and while some may not share his views, few would say that AI will not be adopted at all. Huang believes that the struggle comes from trying to keep up with competitors who press ahead, leveraging AI to implement cost and productivity gains that allow them to steal your customers while you hesitate. Nvidia’s AI-powered future is just beginning, and the company’s decision to embrace AI-powered image processing in 2018 has redefined its future and helped to redefine an evolving industry.