NVIDIA’s Jensen Huang Claims That Moore’s Law Is Now Dead, It’s Time For The Huang Law; Reveals Performance Is On-Track To “Double” In Just a Quarter

May 22, 2025 at 07:12am EDT

NVIDIA has managed to raise the level of AI performance to an extent that wasn't deemed possible, and according to Jensen, Moore's Law had less of a role.

NVIDIA Isn't Stopping With The AI Hype Anytime Soon, Reveals They Defied Moore's Law Through Native Technologies

For those unaware of what Moore's Law actually is, it is more of an observation that states that node shrinking is directly related to the performance of a machine, and the timeline for this law was set for two years. This means that any "incremental" performance upgrade will come after twelve months. And, if you look at what's happening in the AI world, we are getting a new and advanced architecture every six months, thanks to how NVIDIA is operating with the hype to stay relevant. Jensen says Moore's Law isn't appropriate at all in modern times, and that the "sky is the limit".

Related Story “I Produce The Lowest Cost Tokens In The World” Says NVIDIA CEO As He Highlights The Full-Stack Approach To AI

NVIDIA's CEO appeared at Foxconn's keynote at Computex (via Taiwan Economic Daily), where he was asked whether Team Green anticipates a drop in AI chip development or its pace. In response, Jensen said that Moore's Law has become ineffective simply because the physical limitations have increased tremendously over the past few years. In particular, he was referring to how the drop in node size doesn't bring as much of a performance increase as what NVIDIA delivers through other means.

So, the interesting question here is, what has changed? Well, Jensen believes that advanced packaging technology has played a huge role in delivering the performance increase we have seen with architectures, notably the CoWoS technology by TSMC. Moreover, NVIDIA's proprietary interconnect technology, the NVLink, has allowed the firm to scale up chip performance by connecting multiple platforms in rack-scale solutions, allowing them to deliver computational capabilities deemed impossible.

NVIDIA doesn't look to stop just yet, since Jensen did say that the firm could very well adjust its product roadmap to a quarterly basis, which is an achievement that is not only impressive, but deemed impossible for now. Right now, Team Green is operating on a "six-month" product cadence, at least when you look at the release schedules announced by the company. Vera Rubin and Blackwell Ultra are separated by approximately six months, and this shows that the firm isn't stopping with the AI hype just yet.

About the author: Muhammad Zuhair is a hardware and technology reporter for Wccftech, specializing in the semiconductor industry and the complex interplay between technology, manufacturing, and geopolitics. His coverage focuses on the corporate strategies and technological roadmaps of industry giants like TSMC, NVIDIA, Samsung, and Intel. Zuhair's expertise lies in deconstructing complex topics such as fabrication nodes (e.g., 2nm process), the economic impact of policies like the CHIPS Act, and the strategic development of AI infrastructure from NVIDIA, AMD and Intel.

Follow Wccftech on Google to get more of our news coverage in your feeds.