Hua Er Jie Jian Wen·2026-01-28 01:12Core Insights - Nvidia plans to shift part of its chip manufacturing to Intel, collaborating on the Feynman architecture platform expected to launch in 2028, reflecting a broader trend among U.S. tech companies to diversify supply chains amid political and economic pressures [1][2] Group 1: Nvidia and Intel Collaboration - Nvidia will adopt a "low-volume, low-tier, non-core" strategy in its collaboration with Intel, with core GPU chips still being manufactured by TSMC, while I/O chips will utilize Intel's 18A or the anticipated 14A process [1][2] - The collaboration is part of a strategic shift in response to U.S. manufacturing goals and tariff pressures, with other companies like Google, Microsoft, AWS, Qualcomm, Broadcom, AMD, and Tesla also in discussions with Intel [1][2] Group 2: Impact on TSMC - Despite some orders being diverted to Intel, industry analysts believe the impact on TSMC will be more beneficial than detrimental, as it helps alleviate monopoly concerns and political pressures while maintaining confidence in securing high-end chip orders [1][5] - TSMC's strategy includes managing customer shifts by focusing on non-core orders, which may enhance its bargaining power and supply capabilities in the future [5] Group 3: Broader Industry Trends - The shift towards Intel by companies like Nvidia and Apple is driven by the need to mitigate risks associated with single-source manufacturing and to address capacity shortages [3][4] - The collaboration with Intel is seen as a response to the challenges posed by the U.S. manufacturing goals and the need for supply chain resilience [2][3]