Nvidia boss Jensen Huang has discussed the advantages of AI, and said it was now a key component for the advancement of graphical capabilities. "We can't do computer graphics anymore without artificial intelligence," Huang said, speaking recently at the Goldman Sachs Communacopia + Technology conference (via SeekingAlpha ). "We compute one pixel, we infer the other 32.
I mean, it's incredible," Huang continued. "And so we hallucinate, if you will, the other 32, and it looks temporally stable, it looks photorealistic, and the image quality is incredible, the performance is incredible." While the energy costs of using AI are often highlighted as one of its various downsides, Huang argued that using it to compute parts of an image was instead saving power versus creating the image directly.
"The amount of energy we save - computing one pixel takes a lot of energy," Huang added. "That's computation. Inferencing the other 32 takes very little energy, and you can do it incredibly fast.
" Nvidia uses AI generation in its powerful DLSS technology, which boosts performance by AI-generating additional frames. Last week, we learned more about how Sony was implementing a similar idea with PSSR AI upscaling in the freshly-unveiled PS5 Pro . AI is frequently criticised for its planet-melting energy costs, particularly in the growing usage of publicly available large language models (LLMs) and image generation tools such as ChatGPT and Copilot.
Still, Huang said the "productivity gains" from.