Following Apple, NVIDIA also plans to shift part of its chip manufacturing business to Intel's foundry. NVIDIA's Feynman architecture platform, expected to launch in 2028, will collaborate with Intel, marking the latest case of American tech giants adjusting their supply chain strategies under the &#34;Made in America&#34; goal promoted by the Trump administration.
According to a report by DIGITIMES on Wednesday, supply chain insiders revealed that NVIDIA will adopt a &#34;small volume, low-end, non-core&#34; collaboration strategy in its partnership with Intel. The GPU core chips will still be manufactured by TSMC, while the I/O chips will partially use Intel's 18A process or the 14A process scheduled for mass production in 2028, with advanced packaging ultimately done by Intel's EMIB. In terms of advanced packaging share, Intel accounts for about 25%, while TSMC accounts for about 75%.
This shift reflects the necessity for American tech companies to move from a highly concentrated model focused on TSMC to a new strategy of &#34;multi-source supply and risk diversification&#34; under political pressure, tariff threats, and supply chain resilience considerations. In addition to NVIDIA and Apple, companies such as Google, Microsoft, AWS, Qualcomm, Broadcom, AMD, and Tesla are also in talks with Intel for collaboration.
Although some orders are being diverted to Intel, the industry believes that this is &#34;more beneficial than harmful&#34; for TSMC, as it helps alleviate monopoly concerns and political pressure, while TSMC remains confident in securing large orders for high-end chip foundry services.
<h2>NVIDIA's Feynman Architecture Partially Shifts to Intel</h2>
After announcing a $5 billion investment in Intel in September 2025, NVIDIA's latest plan is to collaborate with Intel on the next-generation successor to the Rubin series, the Feynman architecture chip. According to supply chain sources, the GPU core chips will still be manufactured by TSMC, while the I/O chips will partially use Intel's 18A or 14A processes, with the specific choice depending on the subsequent yield of the 14A mass production.
Supply chain insiders indicated that under the establishment of the &#34;Made in America&#34; goal and tariff pressures during the Trump administration, major American chip manufacturers had already discussed collaboration with Intel, but due to the 18A process not meeting customer expectations, the collaboration timing should align with the 14A process scheduled for mass production in 2028. Intel's CEO Pat Gelsinger recently stated that two customers are currently evaluating the specific details of the 14A process.
Compared to the higher risks associated with the introduction of the 14A and 18A processes, most companies are starting their collaboration with Intel through advanced packaging EMIB first. Supply chain analysts noted that the &#34;Made in America&#34; goal faces challenges in cost and yield, but under the realities of political, supply chain resilience, and TSMC's limited advanced packaging capacity, American chip manufacturers are bound to initiate a dual-foundry strategy.
<h2>Apple Restarts Collaboration with Intel for Entry-Level Processors</h2>
The collaboration product that Apple has been negotiating with Intel for some time is likely the &#34;entry-level M series processors&#34; used in MacBooks, which are currently manufactured by TSMC. Apple's Mac series has been using Intel's x86 processors since 2006, and Intel established a dedicated production line for it in its wafer fab in Oregon, USA, in 2005.
In June 2020, Apple officially announced the launch of its Arm architecture chip &#34;Apple Silicon,&#34; with all Mac series models transitioning to self-developed chips two years later. At that time, Apple's main considerations were to expand supply chain control and integrate its ecosystem, with another key factor being the potential impact of Intel's 10-nanometer process delays on the launch of new Mac models
Supply chain operators reveal that the main reason Apple is restarting its collaboration with Intel after three years is still the manufacturing goals and tariff impacts driven by Trump, followed by factors such as cost, diversifying single-source manufacturing risks, and capacity shortages.
<h2>TSMC's Three-Layer Strategy to Address Customer Diversion</h2>
Currently, it is observed that Apple and NVIDIA are gradually adjusting with products that have the lowest outsourcing risks. Other companies negotiating cooperation with Intel include Google, Microsoft, AWS, Qualcomm, Broadcom, AMD, and Tesla, as well as the U.S. government contracts with the highest degree of control. However, whether Intel can meet the demands of technology leaders who are already accustomed to TSMC's model remains uncertain.
For TSMC, it has anticipated that many customers will shift to Intel for wafer production, but analysts believe that the actual &#34;benefits far outweigh the drawbacks,&#34; with at least three strategic considerations: First, it can reduce monopoly and regulatory concerns; second, it can alleviate political pressure from the U.S.; third, the spillover is merely &#34;non-core&#34; orders, which helps with future bargaining and supply.
On one hand, this alleviates concerns about TSMC's excessively high market share in the wafer foundry industry potentially involving antitrust laws; on the other hand, moderately releasing non-core orders can also reduce pressure from the Trump administration's various demands. Furthermore, TSMC remains confident in securing major high-end chip foundry orders from major manufacturers. Perhaps customers who have tried shifting orders to other foundries will miss the advantages of TSMC, which will give TSMC a better position in future bargaining and supply

Supply chain insiders reveal that NVIDIA's Feynman architecture platform, expected to launch in 2028, will collaborate with Intel, adopting a "small volume, low tier, non-core" cooperation strategy. In addition to NVIDIA, companies such as Apple, Google, Microsoft, AWS, Qualcomm, and Broadcom are also in talks with Intel for collaboration. Analysts believe that under the realities of political factors, supply chain resilience, and limited advanced packaging capacity at TSMC, major American chip manufacturers will inevitably initiate a dual-foundry strategy

- Nvidia plans to shift part of its chip manufacturing to Intel as part of its collaboration on the upcoming Feynman architecture platform expected in 2028.  
- This move reflects a broader trend among U.S. tech firms, pivoting from reliance on TSMC towards diversified supply chains amid political pressures and tariff threats.  
- Although some orders will go to Intel, industry experts believe this shift benefits TSMC significantly, as it alleviates anti-monopoly concerns and maintains TSMC’s position in high-end chip manufacturing.

Wallstreetcn

NVIDIA

VanEck Semiconductor ETF

First Trust Nasdaq Food & Semicon

Direxion Daily GOOGL Bull 2X Shares

Direxion Daily AAPL Bull 2X Shares

Direxion Daily NVDA Bull 2X Shares

Alphabet

GraniteShares 2x Long NVDA Daily ETF

XL2CSOPNVDA

XI2CSOPNVDA

YieldMax NVDA Option Income Strategy ETF

Direxion Daily NVDA Bear 1X ETF

T-Rex 2X Long NVIDIA Daily Target ETF

T-Rex 2X Inverse NVIDIA Daily Target ETF

Direxion Semicon Bull 3X

Report: Supply chain news indicates that after Apple, NVIDIA's next-generation GPU will also collaborate with Intel to please Trump