Semiconductor News 20260620

Today’s semiconductor news focuses on the growing demand for AI and the resulting dramatic shifts in manufacturing and design strategies.Of particular note are developments surrounding TSMC’s enhanced 2nm process, N2P. Qualcomm and MediaTek are set to be among the first to adopt it, thereby boosting the competitiveness of the Android camp, while Apple is clearly positioning itself to counter this with its A21 Pro.Meanwhile, while AMD and Intel are strengthening their collaboration on the ACE specification for AI acceleration on the x86 architecture, SK Hynix is moving forward with discussions on HBM supply in the U.S., accelerating the restructuring of supply chains amid geopolitical constraints.Furthermore, in the GPU market, AMD has hinted at potential price increases for its RX 9000 series due to soaring DRAM prices, raising concerns that rising costs will be passed on to consumers.Overall, technological innovations aimed at building AI infrastructure, coupled with companies’ strategic maneuvering to optimize costs and supply chains, are set to significantly reshape the industry landscape.

Main News

1. TSMC’s Enhanced 2nm Process, N2P, to Be Made Available to the Android Camp First; Enabling a Differentiated Manufacturing Strategy for the Apple A21 Series
   TSMC’s enhanced 2nm process, N2P, is expected to be supplied on a priority basis to Qualcomm and MediaTek’s next-generation flagship chips in the second half of this year.This move is intended to secure a performance advantage for the Android camp and position it to compete with Apple. Apple, on the other hand, will adopt a two-pronged strategy: while using the N2P process for next year’s A21 Pro chip, it will continue to use the conventional N2 process for the standard A21.Analysts view this as a meticulous cost-management strategy designed to optimize development and mass-production costs and maintain product profit margins against the backdrop of rising global manufacturing costs and memory supply shortages.
2. AMD and Intel challenge the AI ​​gap with ACE-powered x86 – integrating matrix multiplication engines and low-precision formats into next-generation CPUs.
   The x86 Ecosystem Advisory Group, led by AMD and Intel, has unveiled the latest specifications for “ACE,” an AI acceleration specification for next-generation CPUs. This aims to dramatically enhance the AI ​​capabilities of the x86 architecture by standardizing matrix multiplication engines and low-precision formats, which are essential for AI processing. Combined with existing extensions such as FRED and AVX10, this represents an important step for the x86 camp to maintain compatibility and scalability against competitors such as ARM in the AI ​​era and strengthen its competitiveness in the market.
3. In response to U.S. export restrictions, South Korea and Anthropic form an AI security cooperation alliance.
   As export controls on AI-related technologies in the United States tighten, the South Korean government and domestic industry have formed an “AI Security Cooperation Alliance” with Anthropic, a leading American AI company. This partnership aims to overcome barriers such as supply restrictions for advanced AI models and computing chips, and to jointly establish AI safety and risk management standards. By adopting Anthropic’s advanced AI alignment technology, Korean companies aim to establish a system to continue technological development and global business expansion under international regulatory conditions.
4. AMD May Plan to Raise RX 9000 Prices by Another 10 to 15 Percent as the “RAM Ap ocalypse” Engulfs the GPU Market
   Supply shortages and soaring prices in the DRAM market—the so-called “RAM Apocalypse”—are dealing a severe blow to the GPU market.According to reports, AMD appears to be planning to raise prices for its next-generation Radeon RX 9000 series by an additional 10 to 15 percent.Over the past six months, major GPU manufacturers and board partners have been forced to make repeated price adjustments, and this latest increase highlights a challenging market environment in which rising component costs are directly reflected in the final product price.
5. Trump’s statement: Apple to collaborate with Intel on chip design and manufacturing in the U.S.
   Trump hinted that Apple intends to collaborate with Intel on chip design and manufacturing within the United States. This statement was made in the context of emphasizing the U.S. semiconductor manufacturing reshoring policy, and if realized, it would mark a major turning point where Apple’s design capabilities and Intel’s foundry capabilities would be combined. However, details about the specific contract details and technical feasibility remain unclear at this time, and future official announcements and developments in the partnership between the two companies will be closely watched.
6. Taiwanese semiconductor equipment supplier records broad recovery in May, driven by AI testing demand
   Taiwanese semiconductor equipment manufacturers showed a broad recovery in May sales. The main driver is the expansion of orders for back-end process (packaging and testing) equipment due to the surge in demand for AI data centers. The increasing complexity of AI chips has increased the importance of advanced testing processes, and Taiwanese suppliers are directly benefiting from investments in AI-related infrastructure. This recovery trend suggests that the overall semiconductor industry cycle is upward, driven by AI demand.
7. According to reports, SK Hynix held discussions in the U.S. regarding HBM supply and local investment plans
   It has been reported that SK Hynix held discussions regarding its supply system for HBM (High Bandwidth Memory)—for which demand is surging as next-generation AI memory—as well as its local investment plans in the U.S.For the company, which aims to expand its HBM market share in the AI server market, building a supply chain in the U.S. is essential. While taking geopolitical risks into account, the main focus of these discussions is believed to be deepening collaboration with major U.S. AI chip manufacturers to secure a stable supply chain.
8. As AI Inference Transforms SSD Demand, Kioxia Curbs Capital Expenditures on NAND
   With the widespread adoption of AI inference processing, the demand structure for SSDs in the storage market is undergoing significant changes.However, Kioxia continues to maintain a stance of curbing capital investment in NAND flash memory. This is seen as a prudent business decision aimed at avoiding the risk of price declines caused by excess supply capacity, given that market demand forecasts remain uncertain. While benefiting from AI demand, the company is pursuing a strategy of adjusting production with a focus on profitability.
9. Samsung Foundry to Manufacture Claros Power Management Chips for AI Data Centers
   Samsung Foundry has been commissioned to manufacture Claros power management chips optimized for AI data centers.Since AI servers consume enormous amounts of power, the importance of efficient power management chips is growing. By leveraging its cutting-edge process technology to strengthen the production of high-value-added chips that contribute to improving data center energy efficiency, Samsung aims to diversify revenue streams for its foundry business and differentiate itself from competitors.
10. Interview: Oppstar strengthens collaboration with customers in Japan and South Korea through ASIC design, plans to open office in Taiwan
    Oppstar, a Malaysian ASIC design service company, has announced plans to strengthen collaboration with customers in Japan and South Korea. As demand for AI and automotive chip design increases, the company plans to open an office in Taiwan to strengthen ties with local supply chains. By utilizing Taiwan as a hub for design contracting in the Asian market, we aim to expand our global market share by providing faster and more advanced design solutions to customers in Japan and South Korea.

News List (164 items)

This article is delivered every morning via an automated collection and AI-generated system. Number of items collected: 164 | Last updated: June 20, 2026

Editor’s Note: This article utilizes AI to summarize and organize news content. While every effort has been made to be as accurate as possible, it may contain errors in background explanation or interpretation of causal relationships. Please always check the source article for details and accurate context.