Rumor has it that Nvidia (NVDA.US) suspends process testing, and the Intel (INTC.US) 18A chip foundry dream was a hit

The Zhitong Finance App learned that the stock price of the established US chip giant Intel (INTC.US) fell during pre-market trading in the US stock market on Wednesday, and once plummeted by more than 5%. Earlier, according to two people familiar with the matter, Nvidia (NVDA.US), the “supremacy of AI chips,” did not continue to move forward after actually evaluating how to manufacture one of the AI chips using Intel's 18A advanced process technology, but chose to stop testing the Intel 18A process.

The world's two major AI chip design leaders (known as fabless in the chip industry) — Nvidia, which focuses on AI GPU computing power systems, and Broadcom (AVGO.US), which focuses on AI ASIC computing power technology routes, revealed in the media as early as early March that these two major chip design leaders conducted high-end chip manufacturing tests with Intel.

However, since then, Intel's 18A chip technology has continued to face challenges because the initial customer test results were not optimistic, leading to customer dissatisfaction with Intel's technical level, and it is likely that it will continue to invest all of its chip foundry/manufacturing capacity into TSM.US (TSM.US), Intel's strongest competitor in the chip manufacturing field.

Intel is currently in urgent need of major customers to support the chip foundry business

For Intel, it is essential to master advanced process chip manufacturing technology that is world-leading and extremely suitable for mass production. This established American chip giant didn't even hesitate to buy a more expensive high-NA EUV lithography machine — this is also the world's first High-NA delivered by Asmack, and Intel strives to create the most advanced chip manufacturing technology at 2nm and below, ahead of TSMC and Samsung Electronics.

Intel is now only selling its own PC/server CPUs, and it is already difficult to achieve sustained growth in performance. Although Intel has come out of the “darkest hour” of poor performance with strong server CPU demand, what shareholders want to see is that Intel regains its strong growth curve and becomes a chip foundry giant comparable to TSMC. Therefore, Intel's strategy has shifted to focusing on Intel Foundry (Systemic Chip Foundry) — that is, to contract high-end advanced process chips for others, directly targeting TSMC.

Among them, 18A and more advanced 16A technology can be said to be the key points for Intel's transformation. If 18A can actually match or even surpass TSMC and Samsung in terms of performance/energy efficiency, Intel will have the opportunity to contract advanced process chips for Nvidia, AMD, and super big customers such as AI GPU/ASIC.

Intel's 18A (equivalent to 1.8 nm manufacturing process) is an important milestone in its advanced manufacturing roadmap. It aims to greatly improve the performance and energy efficiency of next-generation chips by adopting the RibbonFET transistor structure and PowerVia back power supply technology, and compete with advanced manufacturing technology from competitors such as TSMC. The 18A process is not only used in Intel's own CPU series products, but is also regarded as a key technology node for attracting customers in its external foundry (i.e. foundry) business.

For Nvidia, suspending testing may be a risk aversion and supply chain strategy choice; it is not the same as completely denying Intel

Earlier in September, Nvidia and Intel could be called the “partnership of the century”. Nvidia CEO Hwang In-hoon announced that they would spend 5 billion US dollars to invest in long-time competitor Intel and jointly develop high-performance chips for PC and AI data centers. After this news came out, Intel's stock price, which had been weak for a long time, changed its previous trend. Since September, Intel's stock price has risen sharply by more than 50%. Nvidia's investment in Intel has undoubtedly made the market more hopeful that Intel is expected to receive massive AI chip orders from Nvidia in the near future.

As the world's largest AI GPU computing power cluster designer, every generation of AI chip products designed by Nvidia has extremely high requirements on the performance, power consumption, and yield of the chip manufacturing process. The suspension of progress means that Nvidia chose to avoid current process risks and continue to rely on its existing mature partners and chip manufacturers (such as TSMC Advanced Manufacturing) that have cooperated for decades to guarantee product delivery plans. This shows that Nvidia still prioritizes stability and performance in chip supply chain arrangements.

“Suspending the advancement of advanced manufacturing processes” generally emphasizes that current chip prototype tests have not met expectations or have not entered the next stage of commercial cooperation negotiations, which is not tantamount to long-term abandonment. Nvidia has also participated in early testing of 18A in the past and may help Intel improve manufacturing details based on future yield.