SK Hynix ADR fell more than 9%, and the long-term lock-in price detracted from profit expectations; Meta Louisiana AI data center investment increased to $50 billion
｜Tuesday, July 14, 2026｜NO.1 adjusts the Mac chip roadmap to accelerate AI layout, Apple's stock price hit a record high intraday local time On July 12, according to media reports, Apple is adjusting the Mac chip roadmap to adapt to the development needs of AI. According to reports, Apple will skip high-end versions such as M6 Pro, M6 Max and M6 Ultra for the first time after launching the M6 basic chip this fall, and directly promote the research and development of the M7 series. The M7 is expected to be released in the first half of 2027, and the M7 Ultra is planned to be launched in 2028.
｜Tuesday, July 14, 2026｜ NO.1 Adjusts Mac chip roadmap to accelerate AI layout, Apple stock price hits record high On July 12, local time, according to media reports, Apple is adjusting its Mac chip roadmap to adapt to AI development needs. According to reports, Apple will skip high-end versions such as M6 Pro, M6 Max and M6 Ultra for the first time after launching the M6 basic chip this fall, and directly promote the development of the M7 series. The M7 is expected to be released in the first half of 2027, and the M7 Ultra is planned to be launched in 2028. In addition, Apple has started research and development of M8 and subsequent chips. The 2028 products are expected to use the 1.4nm process and further enhance AI performance. On July 13, local time, Apple's stock price rose to US$320.53 during intraday trading, a record high, and its market value once reached US$4.71 trillion. As of the close, Apple rose 0.67% to US$317.43 per share. Comment: Apple has accelerated the pace of AI chip research and development, showing that it is continuing to strengthen the layout of self-developed chips and AI infrastructure. NO.2 SK Hynix ADR fell more than 9%, long-term price lock-in dragged down profit expectations On July 14, local time, SK Hynix ADR closed down 9.32%, almost erasing the 13% increase recorded on the first day of listing last Friday, and the stock price fell close to the issue price of US$149. The market is worried that high-bandwidth memory will be affected by long-term supply agreements and prices will rise slower than expected. Analysts predict that the company's latest quarter operating profit may be about 8% lower than market expectations. In addition, according to media reports, SK hynix is studying to invest part of the ADR proceeds in Korean government bonds. However, so far, the company has not officially confirmed it, and international mainstream media has not independently confirmed it. Comment: The long-term demand logic for AI storage has not changed, but short-term valuations have begun to be tested by performance realization and profit expectations. NO.3 Samsung completes Tesla AI5 chip tape-out, 2nm mass production enters preparation stage On July 13, local time, according to media reports, Kim Jung-kun, chief engineer of Samsung Electronics' wafer foundry, said that Tesla's next-generation AI chip AI5 has completed tape-out and is about to enter the preparation stage for large-scale mass production. The chip will use Samsung's 2nm process and be produced at the Taylor, Texas, factory in the United States. The factory is expected to start initial operations at the end of this year and officially enter mass production in 2027. AI5 is also the first ultra-large commercial order undertaken by Samsung for its 2-nanometer process, and is regarded as an important milestone for its foundry business. Analysts believe that as Tesla orders continue to be shipped, Samsung's long-term loss-making wafer foundry business is expected to improve, but the mass production yield rate of the 2-nanometer process remains a key challenge. Comment: The completion of tape-out of Tesla AI5 means that Samsung’s 2-nanometer advanced process has entered a critical stage of commercialization. If mass production is successful, it is not only expected to improve its competitiveness in the advanced wafer foundry market, but also further strengthen the prosperity of the AI chip industry chain. NO.4 Intel invests an additional 5 billion euros in Ireland to expand the production of AI server chips On July 13, local time, Intel announced an additional investment of 5 billion euros in the Leixlip Manufacturing Park in Ireland, focusing on expanding the production capacity of Xeon server processors and strengthening research and development capabilities to meet the growth in demand for AI infrastructure, while promoting the Intel Foundry wafer foundry business layout. Intel said the investment will further enhance the strategic position of the Irish factory in the global advanced manufacturing system and enhance its ability to provide foundry services to external customers. Comment: As AI data centers continue to expand and demand for server CPUs picks up, Intel is competing for the AI infrastructure market by expanding its advanced manufacturing and foundry businesses. NO.5 Meta Louisiana AI data center investment increases to US$50 billion
On July 13, local time, Meta announced that it would increase the investment scale of the Hyperion AI data center in Louisiana, the United States, from approximately US$27 billion to more than US$50 billion. The planned capacity of the project has also been expanded from 2GW to 5GW. It is expected to have 2GW of computing power in 2030 and complete all construction by 2032. Meta said that the project will become the company's largest AI data center to date and will invest more than US$1 billion in local roads, water supply and other infrastructure. Previously, Louisiana had passed a 20-year sales tax exemption policy for data centers to attract Meta investment. Comment: Meta has once again significantly increased investment in AI infrastructure, reflecting that competition among global technology giants for computing resources continues to escalate. (
On July 13, local time, Meta announced that it would increase the investment scale of the Hyperion AI data center in Louisiana, the United States, from approximately US$27 billion to more than US$50 billion. The planned capacity of the project has also been expanded from 2GW to 5GW. It is expected to have 2GW of computing power in 2030 and complete all construction by 2032. Meta said that the project will become the company's largest AI data center to date and will invest more than US$1 billion in local roads, water supply and other infrastructure. Previously, Louisiana had passed a 20-year sales tax exemption policy for data centers to attract Meta investment. Comment: Meta has once again significantly increased investment in AI infrastructure, reflecting that competition among global technology giants for computing resources continues to escalate. (