NextFin

Nvidia Reportedly Taps Intel Foundry for Next-Gen Feynman GPUs in Strategic Supply Chain Pivot

Summarized by NextFin AI
  • Nvidia is outsourcing a portion of its next-gen GPU production to Intel Foundry, focusing on the upcoming 'Feynman' architecture set for 2028.
  • This partnership marks a strategic shift from Nvidia's reliance on TSMC, aligning with U.S. domestic manufacturing goals and diversifying geographical risk.
  • Intel is expected to handle about 25% of advanced packaging using its EMIB technology, which is crucial for Nvidia's multi-chiplet designs.
  • The collaboration could signal a trend where chip designers view foundries as modular service providers, impacting future relationships in the semiconductor industry.

NextFin News - In a move that could redefine the competitive landscape of the semiconductor industry, Nvidia is reportedly preparing to outsource a portion of its next-generation GPU production to Intel Foundry. According to TweakTown, the collaboration is centered on the upcoming "Feynman" architecture, scheduled for a 2028 debut. While TSMC is expected to remain the primary manufacturer for the high-performance GPU dies, Intel will reportedly take on the production of I/O dies and provide advanced packaging services using its cutting-edge 18A and 14A process nodes.

The reported partnership marks a significant departure from Nvidia’s long-standing, near-exclusive reliance on TSMC for its most advanced AI and gaming chips. The deal, which follows a $5 billion investment Nvidia made into Intel in September 2025, is seen as a strategic effort to align with the domestic manufacturing goals of the current U.S. administration. Under U.S. President Trump, there has been intensified pressure on American technology giants to repatriate supply chains and reduce dependence on East Asian fabrication hubs. By utilizing Intel’s domestic facilities, Nvidia not only diversifies its geographical risk but also secures a secondary source of capacity during a period of unprecedented demand for AI hardware.

From a technical perspective, the Feynman architecture represents the successor to the Rubin platform. Industry insiders suggest that Intel will handle approximately 25% of the advanced packaging for these chips, utilizing its Embedded Multi-die Interconnect Bridge (EMIB) technology. According to Wccftech, Intel recently showcased a breakthrough "thick core" glass substrate integrated with EMIB at NEPCON Japan, a technology that enables denser wiring and better thermal management—critical requirements for the massive multi-chiplet configurations Nvidia envisions for 2028. Intel’s 14A process, which is slated for mass production in the same timeframe, is expected to be the primary node for the Feynman I/O components.

The financial implications for Intel are substantial. Despite a consensus "Reduce" rating from many Wall Street analysts following a mixed Q4 earnings report, the Nvidia rumor has provided a much-needed catalyst for Intel’s stock. According to MarketBeat, Intel shares gapped up to $46.60 following news of the potential deal and insider buying by Executive Vice President David Zinsner. For Intel, securing Nvidia as a foundry customer validates its "IDM 2.0" strategy and proves that its foundry services can compete for the world’s most demanding logic contracts, even as it continues to struggle with yields in its traditional PC processor business.

For Nvidia, the move is a masterclass in risk management. CEO Jensen Huang has previously described the relationship with Intel as one of "lovers, not fighters," acknowledging that while the two compete in the data center CPU and GPU markets, their interests align in building a robust U.S. semiconductor ecosystem. By adopting a dual-foundry model, Nvidia minimizes the risk of a single point of failure at TSMC, which has faced capacity constraints and geopolitical scrutiny. Furthermore, by keeping the core GPU die at TSMC while moving the I/O and packaging to Intel, Nvidia maintains its performance lead while satisfying political requirements for "Made in America" components.

Looking ahead, this partnership could signal the beginning of a broader trend where chip designers no longer view foundries as exclusive partners but as modular service providers. If Intel successfully executes the Feynman contract, it could pave the way for other giants like Apple or Qualcomm to follow suit. However, the success of this transition hinges entirely on Intel’s ability to meet its 2028 production milestones for the 14A node and glass substrate technology. As the industry moves toward 2026 and beyond, the boundary between competitor and partner continues to blur, driven by the dual engines of AI demand and national security policy.

Explore more exclusive insights at nextfin.ai.

Insights

What are the fundamental concepts behind Nvidia's Feynman architecture?

What historical factors led Nvidia to collaborate with Intel for GPU production?

What technical principles are involved in the 14A process node used by Intel?

How is the current state of the GPU market impacting Nvidia's strategic decisions?

What user feedback has emerged regarding Nvidia's shift to a dual-foundry model?

What recent updates have been reported about Intel's foundry capabilities?

What are the implications of the U.S. administration's policies on Nvidia's supply chain strategy?

What potential challenges does Intel face in meeting the production milestones for the Feynman architecture?

In what ways does Nvidia's partnership with Intel challenge traditional notions of competition in the chip industry?

What are the long-term impacts of Nvidia's decision to diversify its foundry partnerships?

How does the Feynman architecture compare to previous Nvidia architectures like Rubin?

What financial trends are influencing Intel’s stock performance amidst the Nvidia collaboration rumors?

What are the broader industry trends indicated by Nvidia's strategic pivot to Intel?

What controversies surround the reliance on domestic manufacturing for semiconductor production?

Which competitors might follow Nvidia's lead in adopting a dual-foundry approach?

What risks does Nvidia face with its dependence on TSMC for core GPU die production?

How might the partnership between Nvidia and Intel evolve over the coming years?

What impact does AI demand have on Nvidia's production strategy and partnerships?

How does Nvidia's collaboration with Intel align with national security policies?

Search
NextFinNextFin
NextFin.Al
No Noise, only Signal.
Open App