Report: Supply chain sources say that after Apple, Nvidia’s next-generation GPU will also cooperate with Intel to please Trump

**Source**: Wall Street Insights

**By**: Zhao Ying

Supply chain sources revealed that NVIDIA's Feynman architecture platform, scheduled for launch in 2028, will partner with Intel. NVIDIA will adopt a **"low-volume, low-tier, non-core"** collaboration strategy in this partnership. Besides NVIDIA, companies including Apple, Google, Microsoft, AWS, Qualcomm, and Broadcom are also in talks with Intel for cooperation. Analysts believe that amid political pressures, supply chain resilience concerns, and constraints on TSMC's advanced packaging capacity, major U.S. chipmakers are bound to adopt a dual foundry strategy.

Following Apple, NVIDIA also plans to shift part of its chip manufacturing to Intel's foundry services. The collaboration between NVIDIA's Feynman architecture platform (set for 2028 launch) and Intel is the latest example of U.S. tech giants adjusting their supply chain strategies under the Trump administration's "Made in America" initiative.

According to a Wednesday report by *DIGITIMES*, supply chain insiders stated that NVIDIA will implement a "low-volume, low-tier, non-core" approach in its partnership with Intel. The core GPU chips will still be manufactured by TSMC, while some I/O chips will use Intel's 18A process or the 14A process scheduled for mass production in 2028, with final advanced packaging completed via Intel's EMIB technology. Based on the proportion of advanced packaging, Intel will account for a maximum of about 25%, and TSMC approximately 75%.

This shift reflects that U.S. tech enterprises, driven by political pressures, tariff threats, and supply chain resilience considerations, have to move away from their heavy reliance on TSMC and adopt a new strategy of **"multi-source supply and risk diversification"**. In addition to NVIDIA and Apple, companies such as Google, Microsoft, AWS, Qualcomm, Broadcom, AMD, and Tesla are also negotiating cooperation with Intel.

Although some orders are being diverted to Intel, the industry believes this is **far more beneficial than harmful** to TSMC. It helps alleviate concerns about monopoly, ease political pressure, and TSMC remains confident in securing large orders for high-end chip foundry services.

### NVIDIA’s Feynman Architecture Shifts Part of Production to Intel

After NVIDIA announced a $5 billion investment in Intel in September 2025, its latest plan is to partner with Intel for the Feynman architecture chips—the next-generation successor to the Rubin series. According to supply chain sources, core GPU chips will still be produced by TSMC, while I/O chips will partially adopt Intel's 18A or 14A process, with the specific choice depending on the subsequent yield and mass production status of the 14A process.

Supply chain insiders indicated that under the Trump administration's "Made in America" goals and tariff pressures, major U.S. chipmakers have long been discussing cooperation with Intel. However, due to Intel's 18A process failing to meet customer expectations, the partnership is expected to kick off with the 14A process, which will enter mass production in 2028. Intel CEO Pat Gelsinger recently stated that two customers are currently evaluating the detailed specifications of the 14A process.

Given the high risks associated with the adoption of the 14A and 18A processes, most industry players are prioritizing cooperation with Intel on advanced packaging via EMIB technology first. Supply chain analysts noted that while the "Made in America" initiative faces cost and yield challenges, the realities of political factors, supply chain resilience needs, and constraints on TSMC's advanced packaging capacity mean major U.S. chipmakers are certain to launch a dual foundry strategy.

### Apple Restarts Intel Partnership for Entry-Level Processors

The product that Apple has been negotiating with Intel for contract manufacturing is expected to be the **entry-level M-series processors** used in MacBook devices, which are currently produced by TSMC. Apple's Mac lineup has used Intel's x86 processors since 2006, and Intel even established a dedicated "Apple Group" production line at its Oregon wafer fab in the U.S. specifically for Apple in 2005.

In June 2020, Apple officially announced the launch of its Arm-based "Apple Silicon" chips, and two years later, all Mac models switched to using self-developed chips. At that time, Apple's main considerations were to enhance supply chain control and integrate its ecosystem; another key factor was that delays in Intel's 10nm process threatened the launch of new Mac products.

Supply chain sources disclosed that the main reasons Apple restarted its partnership with Intel after three years are still the Trump-led "Made in America" goals and tariff impacts, followed by factors such as cost control, diversifying single-foundry manufacturing risks, and addressing capacity shortages.

### TSMC’s Three-Tier Strategy to Respond to Customer Diversion

Current observations show that both Apple and NVIDIA are making gradual adjustments by shifting products with the lowest manufacturing risks. Other companies in talks with Intel include Google, Microsoft, AWS, Qualcomm, Broadcom, AMD, and Tesla, as well as U.S. government agencies with long-term large-scale procurement contracts. However, there remains significant uncertainty as to whether Intel can meet the requirements of tech leaders that have long been accustomed to TSMC's manufacturing model.

For TSMC, it has long anticipated that many customers will divert some orders to Intel, but analysts believe this is actually far more advantageous than detrimental, based on at least three strategic considerations:

1.  **Alleviate monopoly and regulatory concerns**;

2.  **Ease political pressure from the U.S.**;

3.  The diverted orders are only "non-core" ones, which will help improve TSMC's bargaining power and supply flexibility in the future.

On one hand, this helps TSMC resolve concerns over potential antitrust scrutiny due to its dominant market share in the wafer foundry industry. On the other hand, appropriately transferring non-core orders also eases the mounting pressure from the Trump administration's various demands. Most importantly, TSMC remains confident in retaining core high-end chip foundry orders from major clients. Perhaps customers who experiment with shifting orders to other foundries will come to appreciate TSMC's strengths, which will further enhance TSMC's bargaining power and supply advantages in the future.

### Risk Warning and Disclaimer

The market is risky, and investment requires caution. This document does not constitute personal investment advice, nor does it take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this document are consistent with their specific circumstances. Any investment made based on this document shall be at the user’s own risk.