Nvidia CEO Jensen Huang Explains Why the Massive AI Spending Wave Actually Makes Perfect Sense: '... All Of It Justified'

Core Viewpoint - Nvidia CEO Jensen Huang argues that current investments in AI are not indicative of a bubble, but rather a significant transformation in computing technology [1][2]. Group 1: Shift in Computing Paradigm - Huang emphasizes that Moore's Law, which historically allowed CPUs to double in performance, has reached its limits [3]. - The demand for computing is increasing rapidly, but traditional CPUs are unable to meet this demand, leading to a shift towards accelerated computing powered by GPUs, a transition Nvidia has been advocating for over two decades [4]. - A notable statistic is that six years ago, CPUs powered 90% of the world's top 500 supercomputers, whereas today, that figure has dropped to less than 15%, with the majority now utilizing GPUs [5]. Group 2: Data Processing and Recommender Systems - The largest computing workloads have transitioned to GPUs, driven by the need for data processing across various sectors such as banking, e-commerce, and advertising, which collectively spend hundreds of billions annually on data processing [6]. - Recommender systems, which Huang describes as "the engine of the internet today," rely heavily on algorithms that operate on GPUs, influencing nearly every digital interaction [7]. Group 3: Future of AI - The emergence of agentic AI, including models from companies like OpenAI and Google, is seen as a subsequent development following the foundational shift towards GPU-based computing [7]. - Huang concludes that the resources available to support this revolutionary agentic AI are significantly less than previously assumed, yet still justified [8].