Intel stock price soars 12%! Google shifts orders to Intel to produce 3 million TPUs, Nvidia is also testing the 18A process

According to a report by The Information dated June 8, 2026, due to tight advanced process and packaging capacity at TSMC, tech giants Google and Nvidia are actively seeking to diversify their supply chains and are considering using Intel as a backup manufacturer for advanced AI processors. Driven by this positive development, Intel (NASDAQ: INTC) shares surged by about 12% in early trading on Monday.
(Background recap: Breaking news—Apple and Intel form a shocking alliance! Rumors say Intel will manufacture chips for Apple, sending the stock soaring 14% to a new all-time high.)
(Additional background: Intel (INTC) stock skyrocketed 11% to a 52-week high! Market cap surpassed $530 billion.)

Table of Contents

Toggle

• Google places an early order, delivering 3 million TPUs by 2028
• Nvidia tests the 18A process, looking at the new Feynman architecture
• The trend toward supply chain diversification takes shape, and Intel finds an opportunity

As the global artificial intelligence (AI) boom continues to run hot, demand for high-end computing power is growing explosively, placing severe strain on TSMC (TSMC)’s advanced process wafer lines and advanced packaging capacity. To ensure supply chain stability, U.S. technology giants are actively looking for a reliable second supplier.

According to a report by The Information dated June 8, 2026, citing four people familiar with the matter, Google and Nvidia are considering adding Intel (NASDAQ: INTC) to their list of backup manufacturers for advanced processors. Once this news broke, it immediately injected confidence into the market, driving Intel’s share price up by about 12% in early trading on Monday.

Google places an early order, delivering 3 million TPUs by 2028

The report said that after several months of testing Intel’s advanced packaging technology, Google has decided to take concrete action. Google has formally placed an order with Intel, and expects Intel to produce more than 3 million tensor processing units (TPUs) for it in 2028.

TPUs are Google’s proprietary AI chips developed in-house, mainly used to train and run massive AI models. According to Morgan Stanley’s prior estimates, Google’s total TPU output between 2027 and 2028 will exceed 6 million units, meaning Intel would take on a very substantial foundry order—becoming a key enabler of Google’s AI infrastructure.

Nvidia tests the 18A process, looking at the new Feynman architecture

On the other hand, although Nvidia—the dominant player in AI chips—has not yet officially placed an order with Intel, technical cooperation between the two is reportedly progressing quietly. It is said that Nvidia is currently testing Intel’s packaging technology to assess whether it can manufacture a high-end processor that “integrates 4 graphics processing units (GPUs) into a single computing unit.”

This test is closely related to Nvidia’s planned rollout of its new “Feynman” series GPU architecture in 2028. In addition, Nvidia has already carried out early experimental wafer runs using the multi-project wafer (Multiproject Wafer) model on Intel’s most advanced 18A process node.

The trend toward supply chain diversification takes shape, and Intel finds an opportunity

With TSMC operating at full capacity, the “backup manufacturer” strategy for major tech firms to diversify supply-chain risk is becoming increasingly apparent. If Intel can successfully seize this wave of order switching and backup demand, and prove the reliability of its advanced process and packaging technologies, it could not only significantly boost foundry business revenue, but also help it regain a foothold in the fiercely competitive wafer foundry market—benefiting from the massive windfall from AI infrastructure.