Everything About AI
NextFin News - On May 6, 2026, Jensen Huang, founder and CEO of NVIDIA, joined a CNBC segment remotely while Corning chairman and CEO Wendell P. Weeks appeared at an investor event at the New York Stock Exchange to discuss NVIDIA’s multiyear partnership with Corning and related industry developments. The companies said the collaboration will dramatically expand U.S. optical connectivity manufacturing capacity and boost U.S. fiber production, and NVIDIA separately announced a strategic pact with IREN to accelerate DSX‑aligned AI deployments. (manufacturingdigital.com)
Jensen Huang’s five‑layer framework for AI infrastructure
Huang described the AI transformation as a layered industrial buildout he often explains as a "five‑layer cake." He laid out the layers in sequence and tied each to specific industry activity and jobs:
The first layer of the AI cake is power.
He argued that large data centers require substantially more electricity and that utilities and power‑generation projects will be central to the buildout. He said sustained demand for energy is returning for the first time in decades and that bringing sustainable energy and grid upgrades online is part of the opportunity.
Next layer are chips and semis.
Huang noted that chips and the machines that make them — plus supporting materials such as copper and fiber — form the next layer and require expanded manufacturing and supply‑chain investment.
Layer three... infrastructure.
This layer covers data‑center construction, builders and cloud providers. Huang emphasized that these activities translate into concrete jobs — electricians, construction workers and skilled craft professionals.
Fourth layer, Jensen calls them models.
He listed model builders and labs such as Google, Meta, OpenAI and Anthropic as part of the layer of models that drive compute demand.
Finally, the last layer applications.
Huang pointed to thousands of applications across software and security firms and said this application layer expands the market and creates further employment.
AI as an infrastructure buildout and jobs creator
Throughout the interview Huang repeatedly framed AI as an infrastructure program, not merely a software trend. He said that AI will be deployed as foundational infrastructure globally and that the market is generating large amounts of investment:
We're going through the single largest infrastructure buildout in human history. Artificial intelligence is going to become fundamental infrastructure all over the world and surely here in the United States.
He connected the buildout to American manufacturing and job creation: We can use these market dynamics to reinvest, revitalize American manufacturing... and we're building a new infrastructure. We're revitalizing American technology and American manufacturing.
He emphasized that each manufacturing job spawns supporting jobs and that demand for skilled trades has risen materially.
NVIDIA–Corning partnership: scale, optics and "Made in America"
Huang and Weeks described the new NVIDIA–Corning agreement as focused on advanced optical connectivity and large‑scale U.S. manufacturing. Huang framed the deal as part of a strategy to bring the supply chain along as NVIDIA scales:
Together with Corning, we are inventing the future of computing with advanced optical technologies — building the foundation for AI infrastructure where intelligence moves at the speed of light while advancing the proud tradition of Made in America.
Wendell Weeks explained the practical goals of the tie‑up: Inventing the future of computing with advanced optical technologies, building the foundation of AI infrastructure where intelligence moves at the speed of light. And finally... building on the tradition of made in America.
He added that NVIDIA’s commitment helps fund the capital required to build new factories and expand U.S. production capacity.
Optical connectivity, silicon photonics and replacing copper
Both executives discussed the role of optical interconnects in next‑generation data centers. Huang said the architecture of future AI nodes will require enormous numbers of high‑speed interconnects and that silicon photonics will be critical:
This is really about inventing a next generation data center where the amount of interconnects, the number of GPUs, the speed of the interconnects and the scale of the computing is just utterly gigantic.
On copper versus fiber, Weeks and Huang described a staged shift: copper remains inside server racks today, but fiber and optical technology will scale dramatically between racks and across large AI nodes. Huang said Corning will increase U.S. optical capacity by a factor of ten and that NVIDIA’s investment includes a multi‑billion dollar prepayment to support factory buildouts.
Platform breadth, competition and compute‑constrained markets
Huang addressed competition and why NVIDIA’s platform remains central to many models and workloads. He characterized NVIDIA’s architecture as broadly versatile across model types and scientific domains:
We're the only AI architecture that runs every AI... because of the versatility of our platform, we're seeing tremendous growth.
He also emphasized that, for the near term, the industry remains compute constrained: as demand for inference and agentic systems accelerates, virtually every chip that can be produced will be sold. Huang referred to multiple model providers joining the NVIDIA platform and to deployments beyond large language models — including physics, chemistry and robotics.
Anthropic, OpenAI and surging enterprise demand
The interview touched on rapid adoption among model providers and enterprise adoption of agentic AI. Huang and the host referenced surging usage and revenue at Frontier AI labs. The interview presented a short clip and comments noting extraordinarily rapid growth, and Huang tied that growth to increased token and compute demand that helps explain industry supply pressures and ramp plans.
NVIDIA–IREN partnership and DSX AI factories
Separately discussed in the segment was NVIDIA’s agreement with IREN to accelerate DSX‑aligned AI infrastructure deployments. Huang summarized the intent and scale of that collaboration:
We intend to support deployment of up to 5 gigawatts of NVIDIA DSX‑aligned AI infrastructure across IREN's global data center pipeline over time.
The interview noted that IREN’s Sweetwater campus in Texas — a planned 2‑gigawatt site — is expected to serve as a flagship DSX deployment, and that the IREN agreement includes a five‑year right for NVIDIA to purchase up to 30 million IREN shares at $70 per share (a potential $2.1 billion investment) subject to conditions.
Execution, timing and Iron/IREN financial details
The conversation also reviewed near‑term execution risks and timing for data‑center commissioning. Huang and the host discussed the commissioning of GPUs and the cadence of revenue recognition, stressing that while some quarters can be backend weighted, the long‑term demand remains strong. The interview included summary points on how partnerships and prepayments help underwrite factory and plant construction.
On geopolitics and a possible trip to China
Asked about reports that CEOs might accompany the U.S. president to China, Huang answered cautiously about any travel plans, saying he would defer to official invitations but would consider it a privilege to represent the U.S. if invited: If invited, it would be a privilege. It would be a great honor to represent the United States and to go to China with the President.
Product roadmap and "physical AI"
Huang and the host discussed NVIDIA’s data‑center roadmap through 2028 and beyond. Huang outlined successive product launches and argued that NVIDIA’s platform approach — including tools such as Omniverse and AGX for on‑device inference — positions the company across training, simulation and real‑world robotics. He described "physical AI" as the next large wave of demand where physical agents perform real‑world tasks and said NVIDIA is embedding itself across that stack.
Below are the source materials and videos referenced in the segment and news reporting about the announcements:
Reference: SDxCentral — "Nvidia continues optical spending spree with multiyear Corning deal". (sdxcentral.com)
Reference: Manufacturing Digital — "NVIDIA & Corning: Manufacturing AI Infrastructure in the US". (manufacturingdigital.com)
Reference: Fierce Network — "Corning to build 3 new optical plants in the U.S. with a little ($500M) help from Nvidia". (fierce-network.com)
Reference: GlobeNewswire / TMCNet — "NVIDIA and IREN Announce Strategic Partnership to Accelerate Deployment of up to 5 Gigawatts of AI Infrastructure". (tmcnet.com)
Explore more exclusive insights at nextfin.ai.
