News

The two companies announced in a joint statement on Monday that Nvidia will purchase bonds convertible into MediaTek stock.   This investment expands the cooperation between two chip design companies. According to the agreement, MediaTek will use Nvidia's NVLink Fusion and newly released NVHBM technology as part of its technology suite to help facilitate smoother communication between data center components. This smaller company is attempting to challenge Broadcom and Marvel Technologies by helping data center owners create their own components. For Nvidia, this new agreement is its latest move to ensure that the future of artificial intelligence is built around its hardware platform and standards. The company has been building data center interconnect systems, such as NVLink Fusion, and collaborating with companies like MediaTek to ensure its core position throughout the entire hardware chain, not just limited to its leading AI accelerators.   Previously, Amazon also announced a similar agreement, agreeing to deploy an additional 2 million Nvidia components. More importantly, Amazon has also promised to use Nvidia's connectivity technology in its self-developed chips.   Nvidia CEO Huang Renxun, born in Taiwan, is strengthening ties with a company committed to reducing its dependence on the smartphone market. Earlier this year, MediaTek partnered with Google, a subsidiary of Alphabet, and gained increasing recognition as a design partner for a large technology company seeking to build its own AI chips. This development momentum has led to a nearly two-fold increase in the market value of this chip manufacturer in recent months. The agreements with MediaTek and Amazon give Nvidia the opportunity to maintain its core position in the construction of artificial intelligence data centers, even as the company deploys chips aimed at replacing its main products. This transaction once again demonstrates Nvidia's investment as an industry leader in customers, competitors, and other AI related enterprises. Despite Nvidia's strong denial that it is creating false demand, there are still concerns among investors about these seemingly cyclical investments.   The management of the company stated that the huge investment is just another way to accelerate the popularization of its technology and will drive the development of the artificial intelligence industry. Nvidia remains a major beneficiary of significant investments by some of the world's largest companies in artificial intelligence data centers. Its acceleration chip derived from graphics processors performs well in creating and running artificial intelligence tools and services. With this advantage, Nvidia is constantly developing other technologies including complete computers and networks.   Nvidia's leading position has attracted numerous potential competitors and prompted its largest customers, including Amazon, Google, and Microsoft, to increase their internal chip research and development efforts. At present, Nvidia's dominant position remains solid: the company predicts a significant increase of 70% in sales next year.   MediaTek has partnered with Nvidia to help this American chip manufacturer enter the laptop market. This Taiwanese company is assisting Nvidia in developing RTX Spark products for this market.   MediaTek also stated that it plans to actively recruit talents to better seize the business opportunities brought by the artificial intelligence boom. The company expects revenue from artificial intelligence chips to reach approximately $2 billion this year and has set a goal of occupying 15% of the $80 billion data center market next year. NVIDIA and MediaTek deepen long-term partnership to jointly build an artificial intelligence platform from edge to cloud computing NVIDIA and MediaTek announced today the deepening of their long-term cooperation to jointly build the next generation of artificial intelligence computing platforms, covering areas such as artificial intelligence infrastructure, local artificial intelligence computing, and automotive applications.   As part of this collaboration expansion, MediaTek will adopt NVIDIA NVLink Fusion ™  The platform provides a pre validation path for ultra large scale data centers, cloud service providers, and cutting-edge model developers to develop customized XPUs and integrate them into NVIDIA NVLink ™  Connected rack mounted artificial intelligence factories.   Both parties will combine NVIDIA's accelerated computing, artificial intelligence, graphics, and software platforms with MediaTek's leading advantages in custom chips, high-performance computing, energy-efficient System on Chip (SoC) design, advanced packaging, interconnect, and connectivity technologies.   NVIDIA also invested $3.5 billion to purchase convertible bonds issued by MediaTek.   NVIDIA and MediaTek are collaborating in three major areas:   Artificial Intelligence Infrastructure: MediaTek will collaborate with NVIDIA's NVLink Fusion ecosystem to help customers develop customized AI infrastructure for integration with NVIDIA rack level systems and AI factories.   Local Artificial Intelligence Computing: Both parties will continue to collaborate on the development of multiple generations of NVIDIA RTX Spark ™  DGX Spark ™  PC chips integrate NVIDIA GPUs with MediaTek SoCs to power consumer PCs, artificial intelligence developer supercomputers, and enterprise workstations.   In the automotive field, MediaTek and NVIDIA will continue to develop software defined automotive platforms driven by artificial intelligence to welcome the arrival of the era of physical artificial intelligence.   Artificial intelligence is changing every computing platform - from the world's largest artificial intelligence factory to personal computers and cars, "said Huang Renxun, founder and CEO of NVIDIA. MediaTek is one of the world's leading semiconductor companies with extraordinary expertise in system on chip design, connectivity, superior performance, and energy efficiency. We will work together to build a platform that brings NVIDIA's accelerated computing to new markets and allows customers to freely create differentiated AI systems on a large scale. ”   MediaTek and NVIDIA share a common vision of popularizing advanced AI computing across the entire technology field, "said Cai Lixing, Vice Chairman and CEO of MediaTek. NVIDIA's investment strengthens our collaboration in the automotive industry in the era of cloud AI infrastructure, local AI computing, and physical AI. By combining NVIDIA's leading position in accelerating computing and AI software ecosystems with MediaTek's expertise in a diverse AI technology portfolio from edge to cloud, and our leading position in custom chips, we can accelerate customer innovation. ”   Building customized AI infrastructure using NVLink Fusion platform   MediaTek will provide the NVLink Fusion platform as the design foundation for custom AI accelerators developed by customers, enabling their platform to develop in sync with future NVIDIA architectures.   The NVLink Fusion platform provides a pre built, pre certified, and system pre validated infrastructure for multi chip XPU development, accelerating the development process from chip and advanced packaging to rack level systems.   The NVLink Fusion platform integrates key technologies for customized XPUs, including:   The NVIDIA NVLink Fusion chipset utilizes NVIDIA photon or electrical interconnect technology to connect XPUs to the NVIDIA NVLink expansion architecture.   NVIDIA NVLink-C2C， Provide high bandwidth, energy-efficient connections between XPU, NVIDIA Rosa CPU, and other compatible processors.   NVIDIA NVHBM， Integrate customized memory functionality to increase bandwidth and energy efficiency while allocating more chip area for computing.   Building custom accelerators is just the beginning of deploying custom XPUs in rack level AI factories. Integrating multi chip architecture, advanced packaging, high-speed SerDes, HBM, I/O, and scalable networks into a manufacturable and mass-produced system requires extensive engineering design, certification, and supply chain support from chip to rack.   Customers do not need to design and certify every component of custom XPU from scratch, but can focus their resources on defining the differentiated computing capabilities of their platform, while relying on NVLink connectivity, memory architecture, packaging, manufacturing, and rack level technologies provided by NVIDIA and MediaTek to achieve production deployment.   Customers can submit their XPU designs to MediaTek and customize connectivity, memory, packaging, performance, and power characteristics based on their workload and infrastructure requirements.   In addition, MediaTek is also part of the broader NVIDIA NVLink Fusion ecosystem, which enables hyperscale data centers, cloud service providers, and cutting-edge model developers to seamlessly connect custom XPUs to NVIDIA's AI infrastructure.   With NVLink Fusion, partners can leverage NVIDIA's mature vertical and horizontal scaling technology stack and ecosystem, as well as NVIDIA MGX ™  Rack level architecture reduces development complexity, improves performance, and accelerates the time to market for semi custom AI factories. Introducing supercomputers into local AI computing   NVIDIA and MediaTek are working together to advance local AI computing to embrace the era of generative and intelligent AI. Both parties will combine NVIDIA's leading position in accelerated computing with MediaTek's expertise in high-performance, energy-efficient system level chip design.   MediaTek has collaborated with NVIDIA to develop the GB10 Grace Blackwell supercomputer, which powers the NVIDIA DGX Spark. The NVIDIA DGX Spark integrates NVIDIA Blackwell GPU and Grace CPU, and is connected through NVLink-C2C to bring powerful AI capabilities to edge systems. The two companies have expanded their collaboration to NVIDIA RTX Spark, aiming to power the next generation of consumer PCs for the AI era.   Advancing the development of AI defined cars in the era of physical AI.   In addition, NVIDIA and MediaTek are collaborating across multiple generations to advance the development of AI driven software defined cars. Both parties will leverage MediaTek's leading position in automotive system level chip design and NVIDIA's expertise in accelerating computing AI、 Combining advantages in graphics and software.   The MediaTek Dimensity automatic platform integrates NVIDIA technology, providing advanced AI and NVIDIA RTX for intelligent car cockpits ™ Graphics processing capability and compatible with NVIDIA DRIVE AGX ™ Collaborative work. The two companies will continue to build scalable architectures for future generations of products based on this foundation to support increasingly intelligent AI vehicles.

  On August 7, SK Hynix announced plans to invest 54 trillion won (approximately 256.63 billion yuan) to expand its chip factories in South Korea. The company stated that it is building new chip factories in Yongin and Cheongju, South Korea, to meet the "growing demand for memory chips in the age of artificial intelligence." Market research firm Omdia predicts that demand for DRAM and NAND will maintain a compound annual growth rate (CAGR) of 19% from last year to 2030. Omdia believes this growth is not a temporary supercycle, but a structural transformation, where memory will transcend its role as a simple component and become a core infrastructure determining the performance of artificial intelligence. SK Hynix pointed out that in the AI ​​era, technological competitiveness alone is insufficient to maintain an advantage; the ability to deliver sufficient quantities of products at the time customers require is the true competitive advantage. This investment decision was made after a thorough assessment of market demand. The company stated that Yongin Y2 is the second of four wafer fabs being built sequentially within the Yongin Semiconductor Cluster, with a total construction area of ​​approximately 341,000 pyeong (approximately 1.13 million square meters). It will be operational as a DRAM production base. The project is scheduled to commence construction in July next year, with the first cleanroom expected to open in June 2029, primarily producing next-generation DRAM products such as HBM. Currently, the Y1 wafer fab construction is progressing smoothly, with the first cleanroom expected to open in February next year. With the accelerated popularization of AI services, NAND flash memory demand is rapidly growing, primarily driven by enterprise-grade solid-state drives (eSSDs). Simultaneously, the demand for key-value caching in AI inference processes is further increasing, coupled with the gradual implementation of agentic and physical AI applications, the application scenarios for NAND flash memory are expected to expand further. SK Hynix stated that the company chose to build its new NAND flash memory production base in Cheongju because it can accelerate the construction of the new wafer fab. The Cheongju campus has already completed the M11, M12, and M15 NAND flash memory production facilities. The new wafer fab can synergize and enhance efficiency with existing production lines. Meanwhile, the site, power, and water infrastructure are also largely in place. The Cheongju M17 wafer fab has a total construction area of ​​approximately 206,000 pyeong (approximately 680,000 square meters), with construction scheduled to begin in February next year and the first cleanroom to open in December 2028. The investment period will continue until April 2031 and will be implemented in phases according to business progress.

On May 17th, it was reported that with only a few days left until the strike began on May 21st, Samsung Electronics' semiconductor production line has initiated a production reduction program ahead of schedule. This strike involves 43286 workers, more than half of the total number of employees in Samsung's semiconductor division (DS department), making it the largest work stoppage since the company's establishment. It is reported that Samsung has entered emergency management mode and gradually reduced production and operation load. In emergency mode, the factory limits the number of new wafers that can be put into the production process, which directly impacts the semiconductor manufacturing process that typically requires 24-hour uninterrupted operation. The strike target production line mainly produces HBM, LPDDR5, and some customized logic chips, which are the core storage components of current AI computing and high-performance devices. Once the semiconductor production line is shut down, the recovery cycle is extremely long. Even if the strike only lasts for 18 days, it will still take two to three weeks to restart the production line and restore stability, which will impact the global supply of storage chips. The expectation of a strike has spread to the spot market. The price of DDR4 8Gb DRAM chips in the Huaqiangbei market in Shenzhen has increased by about 20% this week, and the price of DDR5 RDIMM 64GB modules has risen to $1350. If customers switch to SK Hynix or Micron due to unstable supply, Samsung may face the risk of long-term market share loss. The South Korean government has warned that the losses will be unimaginable and has stated that it is considering all measures, including emergency arbitration.

On May 5, news emerged that starting from the second half of 2025, the global NAND flash memory market has remained tight, with supply-demand imbalances driving prices to surge. This not only propelled the industry into a high-growth cycle but also made the operational strategies of global leading memory suppliers a focal point of market attention.  It is reported that the significant rise in this round of NAND flash memory prices has already driven leading manufacturers to achieve several-fold growth in revenue and profits, ushering in a comprehensive "boom" for the industry. In terms of profitability, the growth momentum of global top-tier memory manufacturers has been particularly remarkable. Among them, South Korea's semiconductor giant Samsung Electronics reported an operating profit of 57.2 trillion KRW in Q1, a staggering 756% year-on-year increase. SK Hynix recorded a net profit of 40.35 trillion KRW during the same period, up 398% year-on-year.  U.S. memory giant SanDisk also delivered impressive performance. In Q3 of its fiscal year 2026 (i.e., Q1 of 2026 calendar year), it achieved revenue of $5.95 billion under Non-GAAP conditions, marking a 97% sequential increase. Its gross margin reached 78.4%, rising by 27.3 percentage points sequentially, while net profit surged by 280% to $3.675 billion. The company also projected that Q4 revenue would reach $7.75 to $8.25 billion, maintaining strong growth momentum.  This explosive profit growth directly translated to capital markets, with global semiconductor sector stocks performing exceptionally well. In South Korea, SK Hynix and Samsung Electronics saw cumulative one-year stock price gains of 683% and 308%, respectively, while their market values surged significantly. In the U.S., the semiconductor sector also attracted attention, with SanDisk's stock price soaring by 3,542% in the past year, Western Digital rising over 880%, and Seagate Technology climbing more than 700%, making it a "star sector" in capital markets.  It is noted that to secure long-term growth, SanDisk has secured long-term agreements with five clients, three of which were signed in Q3 of fiscal year 2026, while the other two are set to be finalized in Q4. Additionally, more clients are actively in negotiations. Notably, the minimum contract revenue from just the three signed clients reached $42 billion, with the five existing agreements totaling $11 billion in financial guarantees. SanDisk's CFO stated that if these clients fail to fulfill procurement obligations, the company would receive corresponding compensation, further ensuring its earnings stability.  Regarding the industry's future outlook, the sector remains cautiously optimistic. Current market consensus suggests that the supply shortage in the NAND flash memory market will persist until at least 2027.

The era of huge profits! SK Hynix offers a bonus of 6.1 million yuan per person!   On April 21st, according to South Korean media reports, SK Hynix employees are expected to receive an average of about 700 million Korean won (about 3.2 million yuan) in bonuses this year, and this number will even approach 6.1 million yuan next year! SK Hynix cancelled the original bonus cap in September last year and agreed to directly use 10% of its annual operating profit as employee performance bonus. Analysts predict that SK Hynix's operating profit for 2026 will be approximately 250 trillion Korean won, with a corresponding distributable bonus pool size of 25 trillion Korean won, evenly distributed among the company's approximately 35000 employees. In February of this year, the company distributed a dividend of approximately 140 million Korean won (approximately 95000 US dollars) per employee. Investment bank Macquarie has given an optimistic forecast that SK Hynix's operating profit will reach 447 trillion Korean won next year, and under the current dividend system, the average employee bonus will exceed 1 billion Korean won. In sharp contrast to SK Hynix's sky high bonus, Samsung's ongoing labor management conflict has intensified. The Samsung labor union demands that 15% of operating profits be distributed as employee dividends. Based on analysts' forecast of 298 trillion Korean won in operating profits this year, the semiconductor department's 77000 employees alone can receive approximately 580 million Korean won (approximately 396000 US dollars) per person.   But Samsung's management only proposed a 10% profit distribution plan consistent with SK Hynix, which was directly rejected by the union.   At present, the union has planned to hold a large-scale rally at the Pingze wafer factory on April 23rd. Samsung Electronics applied to the court last Thursday to prevent the union's so-called "illegal behavior". Samsung Union Chairman Choi Seung ho revealed that about 200 employees have switched to SK Hynix in the past four months.   The huge potential bonuses for SK Hynix and Samsung employees have sparked rare public opposition in South Korea. A netizen posted that companies benefiting from national infrastructure investment and the 20% tax credit policy of the K-chip Act should share profits more widely. It is estimated that the total amount of tax benefits received by the two companies in the past two years is about 20 trillion Korean won (approximately 93.1 billion yuan).  

The exit of TSMC has triggered a series of events that are actively reshaping and strengthening the power gallium nitride (GaN) market landscape in many ways. It accelerates the ongoing structural transformation, drives the entire ecosystem towards new business models and partnerships, and enhances regional diversification.   The most significant change is evident: power gallium nitride is becoming a market driven by IDM. Infineon is gradually phasing out GaN system products (manufactured by TSMC) and transferring production to an internal factory located in Austria (EU). ROHM is obtaining authorization for TSMC's gallium nitride technology to establish its own internal manufacturing capabilities in Japan. Other industrial enterprises are also adopting this strategy. For example, companies such as Innoscience (China) are expanding their own production capacity while establishing partnerships with companies such as ON Semiconductor and STMicroelectronics. Meanwhile, Texas Instruments and Renesas Electronics are utilizing their 8-inch wafer fabs in Japan to support the research and production of gallium nitride. However, this has not slowed down the development speed of wafer foundries; On the contrary, it has accelerated investment in gallium nitride, with new entrants constantly emerging and geographical distribution changing. Navitas is shifting towards PSMC (Taiwan) and is also collaborating with GlobalFoundries (USA). GlobalFoundries is partnering with Onsemi to position itself as an important alternative solution for the US market, utilizing TSMC authorized GaN technology. VIS (Vanguard) has obtained a technology license from TSMC, and its partners include EPC, Renesas Electronics, Qromis, and Shinetsu Technologies. The OEM also provides a second supplier for Integrated Device Manufacturers (IDMs), enhancing the security of the supply chain and giving system manufacturers more peace of mind. They also contribute to innovation, enabling integrated device manufacturers to quickly produce new product prototypes and validate them before transitioning to large-scale internal production. The addition of new entrants further strengthens this trend: Samsung (South Korea) is launching an 8-inch GaN production line using vertical integration, including in-house wafers. DB HiTek (South Korea) and SK Keyfoundry (South Korea) are also actively expanding their market presence. This is a rebalancing of geographical activities, as ecosystems are gradually shifting from a model centered around Taiwan to regional diversification.   All of this is happening against the backdrop of accelerated adoption of gallium nitride technology in artificial intelligence data centers and the automotive industry, with growth rates of approximately 50% and 70% respectively, driving the market to reach around $3 billion by 2030. In short, TSMC's exit is not a subversion, but an acceleration of the development of existing trends: IDM (Integrated Design Manufacturer) is regaining control of the manufacturing process. The foundry is diversifying its operations and reallocating demand. Production is becoming increasingly geographically balanced.

Although the spot price of DRAM has slightly fallen recently, the surge in price of storage chips has already triggered demand suppression. Ultimately, why did MediaTek and Qualcomm cut tens of millions of smartphone chip production capacity? It is reported that two manufacturers are reducing the production of mobile chips, which are the main focus of mid to low end smartphones, due to the continuously rising prices of storage chips, which have begun to suppress terminal demand. The smartphone industry is facing a dual impact: on the one hand, the price of storage chips has skyrocketed in the past year - the core cause being that under the AI boom, a large amount of global DRAM production capacity has been squeezed by HBM high bandwidth memory; On the other hand, the ongoing turmoil in the Middle East and Iran has had a profound impact on the global logistics supply chain. For example, the situation in Iran has brought about an unexpected chain reaction: a global shortage of helium gas. Helium is a key material used in semiconductor manufacturing for low-temperature refrigeration and high-purity cleaning. The northern gas field in Qatar is a core source of helium gas globally, but due to Iran's blockade of the Strait of Hormuz, related supplies are currently unable to circulate normally. In terms of the storage market, the spot price of DDR5 has started to fall from its high point, and the mainstream view is pointing the finger at Google's TurboQuant technology. Industry concerns: This algorithm, which can compress KV cache without loss and greatly improve inference efficiency, will directly weaken the subsequent memory procurement demand as major manufacturers collectively expand production. However, industry analysts still believe that a drop in spot prices does not mean an immediate reversal of the market. Daishin Securities found through channel research that a leading cloud vendor recently purchased DDR4 servers at a transaction price even higher than HBM3e; The fact that old memory can still sell at sky high prices is enough to indicate that the demand driven by AI is still strong. Returning to the core issue: In recent months, the skyrocketing storage prices have clearly suppressed demand, especially impacting the price sensitive domestic smartphone market. Therefore, both MediaTek and Qualcomm have lowered the production pace of 4nm mobile phone chips, which are widely used in mid to low end models. The scale of this round of production reduction is about 20000 to 30000 wafers, equivalent to 15 to 20 million mobile processors. MediaTek's 4nm product line covers the Dimensity 7200, 7300, 8200, 8450, 9000, 9300, 9400 and other series; Qualcomm 4nm models include Snapdragon 4 Gen2, 4s Gen2, Snapdragon 7s Gen2, Snapdragon 8 Gen3, and more.

On April 17th, International Electronic Business News reported that a distributor revealed to the news outlet that AOS (Austrones System Inc.), a power semiconductor and chip supplier, officially issued a price adjustment notice to its global customers earlier this month. According to the distributor, some AOS products will officially increase in price on April 1, 2026. The following is the main content of the price increase letter from AOS, disclosed to International Electronic Business News by the distributor: AOS stated in the price increase letter that despite continued increasing cost pressures in the industry, the company has made every effort to maintain existing prices for as long as possible. However, given the continued rise in raw material, energy, logistics, and infrastructure costs, a price adjustment is now necessary to ensure sustainable operation and long-term supply reliability. At the same time, the continued growth in demand for power semiconductor products has led to tighter supply in certain product segments. To support future growth and maintain supply continuity, AOS is making targeted investments to increase capacity and further enhance the resilience of its overall supply chain. Starting April 1, 2026, AOS will implement price adjustments for some selected products. The AOS sales team will follow up and provide detailed information on product scope and implementation timeline. Several power semiconductor companies have already raised prices in 2026. Since the beginning of 2026, the power semiconductor industry has seen a new wave of price increases, and AOS's price adjustment is just another microcosm of the industry-wide cost transmission. From international giants like Infineon and ADI, to domestic leaders like CR Microelectronics and New Energy, and now AOS, global power semiconductor manufacturers have formed a coordinated price increase pattern, covering core categories such as MOSFETs, IGBTs, and diodes, with increases generally at 10% or higher. This round of price increases is not a short-term market speculation, but rather an inevitable result of comprehensive pressure on cost-side raw materials, wafer foundry, and packaging and testing, coupled with explosive growth in downstream demand from new energy vehicles, photovoltaic energy storage, and data centers, leading to a continuously widening supply-demand gap. Some industry insiders believe that this round of price increases has clear sustainability, and the industry's high prosperity and upward price trend are expected to continue into the second half of 2027. Going forward, it is crucial to focus on the pace of new domestic capacity release and the realization of downstream demand.

IT Home News, March 8th – According to South Korea's *Chosun*, the Samsung Electronics Union Joint Struggle Headquarters will hold a strike vote among all union members from March 9th to 18th. The final result will determine whether a large-scale strike will be launched, a move that has already raised concerns in the market about the production of Samsung's core semiconductor products. Notably, even before the strike decision is finalized, the union has already announced its policy, explicitly stating that it will "impose unfavorable treatment on employees who do not participate in the strike." This move has directly sparked heated debate in the industry regarding a "blacklist," highlighting the intensity of this labor-management struggle. It is understood that the union headquarters consists of three main entities: the Samsung Electronics branch of the non-corporate union, the All-Samsung Electronics Labor Union (All-Samsung Labor), and the Samsung Electronics Labor Union Association. Currently, it has approximately 89,000 members, accounting for more than 60% of all Samsung Electronics employees. This large membership base means that if a strike is launched, it will have a significant impact on the company's operations. Market analysts point out that Samsung's Pyeongtaek plant is a major production base for its core semiconductor products such as high-bandwidth memory (HBM). Currently, the AI ​​boom is causing strong demand and tight capacity in the HBM market. If the strike goes ahead, it will likely affect the production schedule and delivery time of related products. The union headquarters has clearly stated that if the strike vote receives majority support and the right to strike is successfully obtained, subsequent actions will proceed as planned: a first rally will be held in Pyeongtaek on April 23, followed by an 18-day general strike from May 21 to June 7. The core issue of this labor dispute is the rules for distributing excess profit incentives (OPI)—currently, Samsung Electronics distributes OPIs to employees up to 50% of their annual salary, and the total bonuses for each department cannot exceed 20% of its excess profits. The union is explicitly demanding that Samsung follow the example of its competitor SK Hynix and completely eliminate this incentive cap, allowing employees to more fully share in the company's performance bonuses. To ensure the strike proceeds smoothly, the union is taking strong measures to pressure employees to participate. On March 5, Choi Seung-ho, president of the non-corporate union, publicly stated that during the strike, the union's executive committee planned to occupy the Pyeongtaek office building and hold rallies to put pressure on the company's management. He further clarified, "If any employees continue to work for the company during the strike, the union will compile a special list and prioritize these employees in future union-negotiated matters such as forced job transfers and layoffs." In addition, the union announced that it would officially operate a whistleblower reporting center during the strike, offering rewards to whistleblowers who report and support company management, thereby strengthening the constraints and guidance on employees. Background information indicates that Samsung Electronics' performance in 2025 is impressive, particularly in its semiconductor business (DS division), which benefits from strong demand in the HBM and general DRAM markets, leading to a significant increase in employee OPI (Optical Performance Income) payouts. The Mobile Experience (MX) division even reached the 50% cap. Meanwhile, its competitor SK Hynix introduced more attractive incentive policies in 2025, not only removing bonus caps but also including 10% of annual operating profit in the bonus pool and introducing a combined cash and stock incentive model. This has become a key factor in the current pressure exerted by the Samsung union. Currently, Samsung Electronics has not publicly responded to the union's demands or strike plans. The subsequent power struggle between labor and management and the voting results warrant continued attention.

According to reports, Samsung Electronics' semiconductor division (DS) offered Apple a 100% price increase, which was unexpectedly accepted immediately. Semiconductor industry insiders revealed that Samsung's DS division originally aimed to raise Apple's supply price by about 60%, and used a 100% increase as the starting price in the first round of negotiations, intending to leave room for negotiation on the final outcome. Unexpectedly, Apple accepted on the spot. Sources familiar with the matter indicated that this sufficiently illustrates the extent of the competition among mobile phone manufacturers for memory inventory, and has also created a chain reaction. Samsung's semiconductor division subsequently abandoned its long-term supply agreement (LTA) with the Mobile Experience (MX) division, replacing it with quarterly contracts to maximize profits. Against the backdrop of a global shortage of general-purpose memory, to ensure shipments of the Galaxy S26 series, Samsung Electronics' MX division will share the initial mass production of LPDDR5X memory between its own DS division and Micron Technology. Both Samsung's semiconductor division and Micron plan to significantly increase the LPDDR5X supply price in negotiations after the initial mass production, which will have a significant impact on the profitability of Samsung Electronics' Mobile Experience division. Samsung launched its new Galaxy S series flagship models in San Francisco on Wednesday. The base model S26 is priced at $899, and the S26+ at $1099, both higher than their predecessors.

According to South Korean media reports, Samsung Electronics plans to increase prices for process nodes such as 4nm and 8nm, with industry estimates suggesting a price increase of around 10%. On the one hand, the production capacity of mature and sub advanced processes remains tight, with high market demand and limited resources available for expansion. On the other hand, the 4nm and 8nm processes have successfully passed the stage of yield ramp up and process stability, entering the mature mass production cycle. By adjusting prices, the overall profit of mature and high-yield processes is higher.   In terms of specific positioning, Samsung's 4nm process is mainly aimed at customers who have higher requirements for performance and energy efficiency, while the 8nm process focuses more on large-scale production and cost control, suitable for products that are price sensitive but have a large shipment scale. By distinguishing the application scenarios of different nodes and synchronously raising prices, Samsung Electronics hopes to further improve long-term profitability while ensuring stable customer structure. Since entering 2026, multiple global wafer foundries have successively initiated price increases. TSMC has previously stated that it plans to launch a four-year continuous price increase for advanced processes of 5nm and below starting from January 2026, with an average annual increase of about 3% to 5%.   SMIC will increase the price of 8-inch BCD process outsourcing by about 10% in December 2025, and multiple customers have confirmed the relevant adjustments; Huahong Semiconductor, on the other hand, raised the prices of mature processes as early as the second quarter of 2025 and gradually reflected it in subsequent performance. The current increase in wafer foundry prices is still due to the continued surge in demand for artificial intelligence. The demand for AI chips and related accelerators is driving the overall order scale to continuously increase, while advanced processes such as HBM and advanced packaging capacity are being prioritized, further squeezing the available resources of mature and sub advanced processes. In addition, the continuous increase in labor costs, raw material prices, and energy costs has also raised the overall cost of wafer manufacturing.

Tesla CEO Elon Musk recently retweeted a job posting from Tesla's South Korean division, officially announcing the company's efforts to recruit key semiconductor talent in South Korea. In the increasingly fierce global AI chip market, this move is widely interpreted by the industry as a significant strategic move by Tesla to strengthen its semiconductor design and manufacturing capabilities and solidify its competitiveness in the AI ​​chip field. On February 17th local time, Musk retweeted Tesla South Korea's job posting for AI chip design engineers on his social media platform X, actively seeking talent. He stated, "If you want to work in chip design, wafer manufacturing, and AI software related fields in South Korea, you are welcome to apply to Tesla." This talent recruitment plan aligns closely with Tesla's long-term strategy of developing its own chips. Currently, Musk's AI company xAI is fully engaged in developing the Grok large-scale model, directly competing with OpenAI's ChatGPT. High-performance AI chips are the core foundation for supporting the development of large-scale models and achieving efficient processing of massive amounts of data, making Tesla's need for such chips extremely urgent. It is widely believed that Tesla's targeted recruitment drive in South Korea aims to tap into the country's top semiconductor design and process engineers. South Korea boasts a strong foundation in the semiconductor field, particularly in high-bandwidth memory (HBM), a core component of AI chips, where it holds a leading global position. With the explosive growth in global demand for AI computing power, the market demand for HBM is also soaring, making it a key reason for Tesla's focus on the South Korean talent market. Currently, Tesla is collaborating deeply with Samsung Electronics and TSMC to advance the production of AI chips. Samsung Electronics will supply Tesla with the AI5 autopilot chip, which boasts 40 times the performance of its predecessor, designed for the fully autonomous driving system. Reportedly, Tesla initially planned for TSMC to exclusively manufacture this chip, but later adjusted to a joint production model with Samsung Electronics (using a 2nm process), further deepening its ties with South Korean semiconductor companies.