Semiconductor News 20260129

Main News

1. SK Hynix Reverses Samsung with Record Earnings, Invests $10 Billion in the U.S.
   SK Hynix’s 2025 financial results surpassed Samsung Electronics with record operating income of 47.2 trillion won, thanks to an explosion in demand for AI memory. In addition, the company plans to invest $10 billion to establish a U.S. subsidiary specializing in AI solutions to strengthen its AI competitiveness in Indiana and elsewhere in the U.S. This is a strategic move to significantly expand its presence in the U.S. while consolidating its leadership in the HBM market.
2. NVIDIA Considering Outsourcing to Intel for 2028 Chip Production
   NVIDIA is reportedly considering a strategy to transfer some of its 2028 chip production to Intel and reduce its reliance on TSMC, a move that takes into account TSMC’s packaging capacity constraints and geopolitical risk diversification If realized, this would be a major step forward for Intel in the foundry market. If realized, this could have a major impact on the foundry market, and would be an extremely important turning point for Intel in its efforts to revive its manufacturing division.
3. Samsung to Fight Back by Concentrating DRAM Production Capacity for HBM4 Mass Production
   Samsung Electronics plans to allocate most of its latest 1c-nanometer DRAM production capacity to the next-generation HBM4 in an effort to regain its share of the AI memory market lost to SK Hynix, with the goal of starting mass production in early 2026, The company aims to deliver to NVIDIA as soon as possible. In addition, as SK Hynix solidifies its exclusive supply of HBM3E to Microsoft, Samsung is betting its corporate fortunes on a turnaround with the next-generation technology.
4. ASML Plans 1,700 Job Cuts to Boost Efficiency Despite Surge in Orders
   Dutch manufacturing equipment giant ASML has posted record 2025 earnings as orders exceeded expectations on the back of surging AI and memory demand. At the same time, however, the company is planning to cut 1,700 jobs in an effort to make its organization more efficient: demand for EUV lithography equipment is strong, but the company has shown a willingness to review its cost structure and muscle up its organization amid strong earnings to ensure sustainable long-term growth.
5. Beijing Authorities Approve Shipment of NVIDIA H200 to Chinese IT Giants
   Authorities in Beijing, China, have approved the purchase and shipment of NVIDIA’s H200 high-performance AI chips to domestic IT giants such as ByteDance, Alibaba and Tencent. The move, which comes amid ongoing U.S. export restrictions, signals both the reality that China still needs NVIDIA chips for AI development and a pragmatic play in the technology hegemony battle between the U.S. and China. Competition for AI development in the Chinese market may be re-accelerating.
6. Toyota, Suzuki to Adopt Chinese SoCs for In-Vehicle Semiconductors
   Japanese automakers such as Toyota Motor Corporation and Suzuki Motor Corporation are reportedly shifting their strategy to use Chinese-made SoCs (system-on-chips) for their in-vehicle semiconductors. It appears that they have decided that it is essential to introduce locally manufactured semiconductors with superior cost competitiveness and performance in order to maintain competitiveness in the Chinese market, which is leading the way in electric vehicles (EV) and automated driving technology. This event symbolizes the structural change in the supply chain of the Japanese automotive industry.
7. Government has high hopes for NTT’s photoelectric fusion technology, eyeing collaboration with Rapidus
   The Japanese government has high hopes for NTT’s photoelectric fusion (IOWN) technology, which combines electricity and light, and NVIDIA estimates that implementing photoelectric fusion could reduce power consumption by 70%. This will be the trump card for solving the problems of AI data centers. The government hopes to use this technology in conjunction with Rapidus, which aims to mass produce cutting-edge semiconductors, as a catalyst to boost the international competitiveness of Japan’s semiconductor and infrastructure industries.
8. OpenAI’s “200 trillion yen” investment plan for survival
   OpenAI is reportedly envisioning a huge investment plan totaling 200 trillion yen, also known as the “AI version of the Manhattan Project”. This is aimed at building AI infrastructure and securing semiconductors, with the aim of maintaining technological superiority and survival by investing funds that outstrip those of other companies. This is a move that echoes the “Stargate” project in the U.S., and the competition for AI development is entering a capital war at the national budget level.
9. Elon Musk Reduces AI Chip Design Cycle to 9 Months
   Elon Musk has announced plans to significantly shorten the design cycle for AI chips at Tesla and xAI from the previous period to every 9 months. increasing the frequency of hardware updates to keep pace with the speed of AI technology evolution. This is aimed at increasing the frequency of hardware updates in order to keep pace with the evolution of AI technology. This is expected to further accelerate the pace of innovation in the AI semiconductor industry as the company seeks to compete with competitors such as NVIDIA in terms of development speed.
10. Soaring Memory Prices Sharply Shift Prices to End Products, Including “Galaxy S26”
    With the soaring prices of DRAM and NAND flash, it is inevitable that prices will be shifted to smartphones and PCs for the 2026 model year. In particular, there are concerns that high-end models such as the Galaxy S26 will face a significant cost increase, coupled with an increase in memory capacity due to the implementation of AI functions. While there are concerns about the impact on consumers’ willingness to purchase, “chipflation” is underway, which provides an opportunity for memory manufacturers to expand their earnings.

News List