NVIDIA CEO Jensen Huang: x86 and ARM to Coexist in Data Centers

NVIDIA and Intel Forge Landmark $5 Billion AI and CPU Partnership.NVIDIA and Intel have announced a major strategic collaboration. The deal focuses on next-generation data center and consumer computing technologies. NVIDIA will also invest $5 billion in Intel’s common stock.This partnership aims to bridge a critical gap in high-performance AI computing. The agreement was confirmed by both companies’ chief executives. It signals a significant shift in semiconductor industry alliances.

Technical Breakthroughs and Market Expansion

The collaboration centers on integrating NVIDIA’s NVLink technology into Intel’s Xeon CPUs. This move addresses a key limitation in current x86-based AI systems. NVIDIA CEO Jensen Huang explained the technical hurdle during a joint webcast.He stated that NVIDIA’s ARM-based Grace CPU can power massive NVL72 AI rack configurations. Current x86 servers are limited to smaller NVL8 setups due to PCIe interface constraints. Integrating NVLink directly into Intel’s silicon will unlock new scalability for x86 servers.This opens a massive new market for both companies. NVIDIA estimates the opportunity at roughly $30 billion. The partnership will allow them to offer customers a choice between ARM and x86 architectures for large-scale AI.

Foundry Dynamics and Mutual Benefits

A key part of the deal involves NVIDIA becoming a primary consumer of Intel’s data center CPUs. In return, Intel will use NVIDIA’s RTX GPU chiplets for a new consumer PC System-on-Chip (SoC). This ensures the partnership is a two-way street with deep technical cooperation.Questions arose about using Intel Foundry Services (IFS) for NVIDIA manufacturing. Huang acknowledged ongoing work with IFS but emphasized TSMC’s current role. He called TSMC a “world-class foundry” and noted both companies are its successful customers.Intel’s advanced packaging technologies like Foveros may play a crucial role. They could be used to combine the upcoming PC chip’s CPU and GPU chiplets. For now, TSMC remains the primary foundry partner for NVIDIA’s leading-edge process nodes.

NYT Strands Hints and Answers for September 19, 2025

This NVIDIA Intel partnership fundamentally reshapes competitive dynamics in the chip industry. It combines forces to tackle the booming AI market from multiple angles. The deal’s success hinges on seamless execution of these complex technical integrations.

Info at your fingertips

What is the main goal of the NVIDIA Intel deal?

The primary goal is to integrate NVIDIA’s NVLink technology into Intel’s data center CPUs. This will enable much larger AI supercomputers using x86 architecture. It also includes collaboration on a consumer PC chip.

How much did NVIDIA invest in Intel?

NVIDIA is investing $5 billion in Intel’s common stock. This financial commitment solidifies the strategic nature of the partnership. It is a sign of strong belief in the collaboration’s success.

Will NVIDIA use Intel’s foundries to make its chips?

NVIDIA’s CEO praised TSMC and called it a vital partner. While work with Intel Foundry Services continues, TSMC remains the primary foundry for NVIDIA’s advanced chips for the foreseeable future.

What does this mean for the AI computing market?

The deal gives customers a choice between ARM and x86 for large-scale AI systems. It taps into a large, previously unaddressed market. It could accelerate the development of more powerful AI infrastructure.

How will this partnership affect consumers?

Consumers may see new PC processors that combine Intel CPU cores with NVIDIA RTX GPU technology. This could lead to more powerful and efficient laptops and desktops in the coming years.