Journalist

김혜준
Candice Kim, Lim Jaeho
  • SK hynix secures top vehicle safety certification for LPDDR5X memory
    SK hynix secures top vehicle safety certification for LPDDR5X memory SEOUL, January 19 (AJP) - SK hynix said Monday it has obtained the highest functional safety certification for its LPDDR5X automotive memory, as demand grows for reliable semiconductors used in advanced driver-assistance and autonomous-driving systems. The South Korean chipmaker said its LPDDR5X DRAM has been certified at ASIL-D, the highest safety level under the ISO 26262 international automotive functional safety standard. ASIL-D applies to systems directly linked to passenger safety, including autonomous-driving control and advanced driver-assistance systems (ADAS). Certification was conducted by global safety testing body TÜV SÜD, which assessed the company’s development, design, verification and quality-management processes. SK hynix said the certified memory is designed for use in ADAS, autonomous driving and in-vehicle infotainment systems, where stable processing of large volumes of data is required. As vehicles increasingly rely on software-based architectures, known as software-defined vehicles (SDVs), functional safety has become a key requirement for automotive semiconductors, alongside performance and power efficiency. Under ISO 26262, automotive components are evaluated based on severity, exposure and controllability of potential failures, with ASIL-D representing the strictest criteria. SK hynix said its LPDDR5X automotive memory incorporates safety mechanisms including error detection and correction, fault-diagnosis functions and redundancy designs to reduce the risk of random hardware failures. The company said the certification supports its strategy to expand its presence in the automotive memory market, where safety validation is becoming a prerequisite for supplier selection by global automakers. SK hynix did not disclose shipment volumes or customer details. 2026-01-19 17:48:47
  • With tariffs looming, Korea chipmakers run out of room to maneuver in the U.S.
    With tariffs looming, Korea chipmakers run out of room to maneuver in the U.S. SEOUL, January 19 (AJP) - To match the tariff concessions Taiwan has secured in the United States, South Korean chipmakers may be pressed to outline additional U.S. capacity expansion plans. Their options are however constrained as the foundry market faces oversupply risks while large-scale memory fabs pose significant economic and strategic hurdles. U.S. officials have adopted a more hardline tone in recent weeks, openly tying tariff relief to domestic semiconductor investment. Yet experts and chipmakers caution that further U.S. manufacturing commitments are far more difficult for Korean firms than for their Taiwanese counterparts. Following a groundbreaking ceremony for a new Micron Technology plant near Syracuse, New York, U.S. Commerce Secretary Howard Lutnick warned that tariff measures outlined in Washington’s recent trade accord with Taiwan could also affect chipmakers from South Korea. “Everyone who wants to build memory has two choices: They can pay a 100% tariff, or they can build in America,” Lutnick said Friday. Under the Taiwan deal unveiled Thursday, companies expanding U.S. operations will be allowed to import semiconductors tariff-free up to 2.5 times their existing capacity during construction, with reduced tariffs applied above that threshold. Once new facilities are completed, the cap will be lowered to 1.5 times current capacity. The agreement also sets a 15 percent tariff on goods from Taiwan and commits the island’s technology sector to at least $250 billion in direct investment in the U.S. Taiwan Semiconductor Manufacturing Co. (TSMC) is expected to build at least four additional chipmaking plants on top of six already planned, requiring roughly $100 billion in new capital. The Taiwanese government has said it will provide $250 billion in credit guarantees to support the investment. South Korea’s accord with the United States revealed last October, sets a 15 percent tariff on most goods while exempting semiconductor imports for now. The agreement includes a $350 billion South Korean fund for U.S. investments, but $150 billion of that is earmarked for shipbuilding, and the remaining $200 billion is government-led and not limited to semiconductors. Private-sector commitments are far smaller. Samsung Electronics has expanded its planned investment in Taylor, Texas, to $37 billion through 2030, up from an initial $17 billion. SK hynix has announced a $3.87 billion investment in Indiana to build an advanced packaging facility for artificial intelligence memory. How much tariff relief Washington will grant in exchange for those investments is emerging as a central negotiating issue. While tariffs will be imposed only after country-by-country talks are completed, pressure is already building. Washington has increasingly blurred the line between trade and industrial policy, explicitly using tariffs to steer semiconductor supply chains toward domestic production. “In practice, the two are closely intertwined,” said Ahn Ki-hyun, secretary general of the Korea Semiconductor Industry Association. “When the U.S. talks about tariffs, what it really wants is investment. And investment naturally connects to supply chains.” The fundamental difference, industry officials say, lies in industrial structure. “Korea and Taiwan are different in what they make,” Ahn said. “Taiwan is centered on system semiconductors and foundries, while Korea is focused on memory. From the U.S. perspective, the logic is simple — if chips are made in America, the problem is solved.” TSMC has already begun production at its first U.S. fab and is preparing additional facilities. Korea, however, currently has no memory wafer production operating in the United States, a gap that analysts say lies at the core of the challenge. Building memory fabs overseas is technically feasible but significantly more complex. DRAM and high-bandwidth memory production requires large-scale infrastructure, long stabilization periods and extremely high yield control. “Memory fabs are not fundamentally different from system semiconductor plants in terms of construction,” Ahn said. “But unlike foundries, there is currently no memory production running in the U.S., which means the timeline is much longer.” Analysts warn that if Washington benchmarks Korea against Taiwan’s investment framework, Korean chipmakers would face a wide gap between existing commitments and the level of spending needed to secure comparable tariff treatment. New memory fabs would require tens of billions of dollars and several years before meaningful output could be achieved. Samsung Electronics said it is closely monitoring developments. “We are reviewing all news and policy changes related to tariffs,” the company said. “Internal discussions are ongoing, but at this stage, there is no official comment we can provide.” Kim Yang-paeng, a senior research fellow at the Korea Institute for Industrial Economics and Trade (KIET), said that estimating the scale of additional investment required under a Taiwan-style framework would be extremely difficult. “If the United States were to demand a similar investment model from Korea, it is not a matter of tens or even hundreds of trillions of won,” Kim said. “The scale would be virtually impossible to quantify in meaningful terms.” 2026-01-19 16:56:30
  • Korean memory makers share boom year with employees in cash and stock rewards
    Korean memory makers share boom year with employees in cash and stock rewards SEOUL, January 16 (AJP) -South Korea’s two largest memory chipmakers, Samsung Electronics and SK hynix are rolling out handsome cash and stock rewards for their employees amid red-hot earnings and extended stock rally. Samsung Electronics said its Device Solutions (DS) division, which oversees the company’s semiconductor operations, has set its excess profit incentive (OPI) at 47 percent of annual salary, according to an internal notice circulated this week. The rate applies across memory, System LSI and foundry businesses. Based on Samsung’s average employee compensation of 130 million won as of March last year, the average bonus payout is projected at roughly 61 million won per employee. OPI is paid once a year when a business unit exceeds performance targets set at the beginning of the year. The incentive can reach up to 50 percent of annual salary, within a cap of 20 percent of excess profit, and together with the target achievement incentive (TAI) forms the backbone of Samsung’s performance-based compensation system. The DS division’s payout reflects a sharp turnaround in semiconductor profitability. After entering recovery in the second half of last year, shipments rose rapidly across memory products, including high-bandwidth memory (HBM) and commodity DRAM. In the fourth quarter, Samsung Electronics posted best-ever three-month operating profit of 20 trillion won. The chip division is projected to have contributed 80 percent of the profit. The contrast with recent years is pronounced. The DS division’s OPI was set at 14 percent last year amid weak chip conditions and zero in 2023. Within Samsung’s Device Experience (DX) division, the Mobile Experience (MX) business set its OPI rate at 50 percent, translating into an average payout of about 65 million won per employee. The higher rate reflects strong sales of the Galaxy S25 series and Galaxy Z Fold/Flip7, up from 44 percent in 2024. OPI rates for the Visual Display (VD), Digital Appliances (DA), network and medical device businesses were set at 12 percent. Management support, Harman, the win-win cooperation center and the global CS center received 39 percent, while the production technology research institute was assigned 36 percent. SK hynix will again operate a “shareholder participation program” this year, allowing employees to receive part of their annual profit-sharing (PS) bonus in company shares, following the program’s initial rollout last year. According to industry officials on Thursday, the company announced the plan in an internal notice the previous day, with applications open through Jan. 22. Under the program, employees can elect to receive between 10 percent and 50 percent of their PS in SK hynix shares, in 10-percentage-point increments. PS is paid once a year based on annual performance, with a ceiling of 50 percent of annual salary, equivalent to 1,000 percent of base pay. For example, an employee earning 100 million won who opts to receive 50 percent of PS in shares would receive stock worth 50 million won. If the shares are held for one year, the employee receives an additional cash payment equal to 15 percent of the stock value, or 7.5 million won. Unlike last year — when shares could be distributed in four installments in the fourth quarter based on prevailing stock prices — participants this year must receive all elected shares at once at the start of the year. PS payments are expected as early as late January or early February. Under a revised labor-management framework, SK hynix scrapped its previous PS cap and now funds bonuses using the full 10 percent of the prior year’s operating profit. Of the calculated PS, 80 percent is paid in the current year, with the remaining 20 percent deferred over two years in 10-percent installments. Future operation of the stock-based program could be influenced by a proposed third revision of the Commercial Act, which would require companies to retire treasury shares. The bill is scheduled for review at a National Assembly subcommittee meeting on Jan. 21, raising the possibility of legislative action later this month or in March. Industry estimates widely put SK hynix’s operating profit for last year at nearly 45 trillion won. On a simple calculation, that level of profitability would translate into profit-sharing bonuses of roughly 136 million won per employee. SK hynix releases fourth-quarter and 2025 results on Jan. 29. Samsung Electronics shares closed Friday 3.5 percent up at fresh record high of 148,900 won and SK hynix also revisited the peak of 756,000 won. 2026-01-16 15:45:28
  • Memory chips take center stage in AI-driven semiconductor boom
    Memory chips take center stage in AI-driven semiconductor boom SEOUL, January 15 (AJP) - The rapid expansion of artificial intelligence (AI) is reshaping the global semiconductor market, propelling memory chips, once considered a cyclical commodity, to the forefront of the industry's future growth engine. Amid a heated race for data centers to build AI servers and inference systems, demand for high-bandwidth memory (HBM), server-grade DRAM, and storage has surged, creating what analysts describe as a rare "triple supercycle" in the market. U.S.-based Bank of America forecast the global HBM market to grow 58 percent this year to $54.6 billion, while global investment bank Nomura Securities expected the broader memory sector to nearly double to $445 billion in 2026, which illustrates how AI is reshaping the industry. Unlike previous upcycles driven mainly by smartphones and PCs, the current boom is being powered by investment in long-term infrastructure. Goldman Sachs estimates that custom AI accelerators will account for roughly one-third of total HBM demand this year, reflecting a shift from training-led growth to inference-driven expansion. On the supply side, manufacturers are struggling to keep pace. Market researcher IDC projected global DRAM bit supply growth of just 16 percent this year, well below the historical average of around 20 percent as chipmakers divert wafer capacity toward higher-margin HBM products. Another researcher Counterpoint Research projected that prices for high-capacity enterprise-grade 64GB RDIMM modules could jump from about $255 in late 2025 to as high as $700 by Marth this year. The imbalance is accelerating a broader realignment in the industry, long dominated by foundries and logic chips. As AI workloads intensify, the industry's center of gravity is shifting toward memory and advanced packaging — a transition that is bolstering the strategic position of South Korea's Samsung Electronics and SK hynix. As AI accelerates, the industry's focus is shifting toward memory and advanced packaging, bolstering South Korean chipmakers Samsung Electronics and SK hynix. "AI processors such as GPUs require ultra-fast memory to avoid performance bottlenecks," said Ahn Ki-hyun, secretary-general of the Korea Semiconductor Industry Association. "To support large language models, the memory attached to GPUs must deliver the highest possible bandwidth, and at the moment HBM is effectively the only option that meets those requirements." SK hynix, which accounts for about 60 percent of the global HBM market, is set to begin commercial production at its new M15X fab in Cheongju, North Chungcheong Province, starting in February, supplying HBM4 for Nvidia's next-generation Rubin platform. Rival Samsung Electronics is aso expanding sixth-generation (1c) DRAM output at its Pyeongtaek campus, aiming to raise advanced-node DRAM to about one-third of total production by the end of 2026. Industry insiders say competition in HBM is driven less by fundamental technology gaps than by speed. "SK hynix gained a competitive edge by being first to commercialize HBM3E," Ahn said. The current upcycle is also being recognized by the financial community. NICE Ratings said in a recent outlook that the rally in memory chips should be seen as a structural shift rather than a temporary rebound, as sustained AI demand collides with limited near-term capacity expansion, a combination likely to support pricing power and profitability through 2026. But risks remain as rising memory prices are expected to lift smartphone and PC prices by 15 percent to 20 percent this year, according to IDC, while power constraints at data centers could slow the pace of expansion. In China, ChangXin Memory Technologies (CXMT) has expanded its share of the DRAM market to about 6 percent and is stepping up efforts in advanced memory development. But Ahn sees its impact as minimal, likely to be felt more in the mid-to-long term than immediately. "CXMT is unlikely to affect the market in the short term," he said. "But if it eventually reaches comparable levels, price competition will become inevitable." For now, the momentum, however, appears firmly on the side of memory chip makers. What began as an AI-driven surge in processors is rapidly evolving into a broader transformation of the semiconductor value chain, in which memory and packaging are no longer supporting players but have become core growth engines. 2026-01-15 17:50:15
  • SK hynix determined to defend lead through talent and HBM investment
    SK hynix determined to defend lead through talent and HBM investment SEOUL, January 14 (AJP) - Over the past year, South Korea’s memory chipmaker SK hynix has nearly quadrupled its market value and overtaken Intel in sales to become the world’s third-largest semiconductor company after behind Nvidia and Samsung Electronics. The turning point came in 2025, a decisive year for the AI-driven chip upcycle. Global semiconductor sales jumped 21 percent to $793 billion, led by AI processors, high-bandwidth memory (HBM) and networking chips, according to preliminary data released by Gartner on Tuesday. “AI semiconductors — including processors, high-bandwidth memory and networking components — continued to drive unprecedented growth in the semiconductor market, accounting for nearly one-third of total sales in 2025,” said Rajeev Rajput, senior principal analyst at Gartner. The momentum is set to intensify. “AI infrastructure spending is forecast to surpass $1.3 trillion in 2026,” he added. HBM propels SK hynix past Intel SK hynix climbed to third place as its global revenue surged 37 percent on year to about $61 billion in 2025, powered by explosive demand for HBM used in Nvidia’s AI accelerators. Intel slipped to fourth, with its market share falling to around 6 percent — roughly half the level it commanded in 2021. Kim Duk-ki, a professor of semiconductor systems engineering at Sejong University, said Intel’s long-standing dominance has steadily eroded. “Intel once sat firmly on the throne, but its competitiveness has weakened in recent years,” Kim said. “Even with strong backing from the U.S. government and renewed efforts to rebuild its foundry business, its long-term direction remains uncertain.” By contrast, he said, SK hynix’s rise reflects more than a cyclical rebound. “Given the structural shift driven by AI, this should not be viewed as a temporary phenomenon,” Kim said. “The gap between SK hynix and Intel could persist for some time.” From memory supplier to “full-stack AI” player To cement its lead, SK hynix has decided to invest 19 trillion won ($14 billion) to build an advanced AI memory packaging plant in Cheongju, North Chungcheong Province. HBM lies at the heart of the strategy. The technology stacks multiple DRAM layers and integrates them with graphics processing units (GPUs) into a single package. Taiwan Semiconductor Manufacturing Co. (TSMC) currently leads advanced packaging, combining SK hynix’s HBM with Nvidia GPUs using its proprietary Chip-on-Wafer-on-Substrate (CoWoS) process — now a standard for high-end AI chips. SK hynix’s decision to build large-scale packaging facilities is widely seen as an attempt to move beyond the role of a component supplier and realize its vision of becoming a “full-stack AI memory creator.” Its local competitor Samsung Electronics has been pitching its turnkey solution that combines foundry manufacturing, HBM supply and advanced packaging under one roof. The investment is also set to elevate the strategic value of SK hynix’s Cheongju campus as an AI semiconductor production hub. The site already hosts NAND fabs M11, M12 and M15, as well as the P&T3 facility handling back-end processes. To secure next-generation DRAM capacity, SK hynix has committed nearly 20 trillion won to its M15X fab. Originally scheduled for later completion, the facility opened its cleanroom last October and is now installing equipment, with mass production expected to begin as early as the second half of this year. Talent war intensifies As SK hynix climbs the global rankings, it is also investing heavily to secure future growth by retaining and attracting talent. The AI-driven memory supercycle has translated into record performance bonuses, intensifying competition across South Korea’s semiconductor industry. SK hynix recently revamped its profit-sharing system, linking employee bonuses more directly to operating profit and removing caps that previously limited payouts. Industry estimates suggest that if current earnings momentum holds, average bonuses could exceed 100 million won ($75,000) per employee this year, with some projections pointing even higher next year if profits continue to surge. The widening compensation gap is already being felt at rival Samsung Electronics, where employees in the memory division have raised concerns over disparities in rewards despite sharing in the same industry upcycle. The issue has added to labor tensions and renewed calls for changes to performance-based pay structures. The effects are visible upstream as well. Competition ratios for semiconductor majors at universities with recruitment agreements with SK hynix — including Hanyang, Chung-Ang, Sogang and Korea University — have climbed to between 9:1 and 11.8:1, sharply above the average 5.31:1 ratio across 11 elite Seoul-based universities. With AI processors generating more than $200 billion in sales last year and HBM accounting for nearly a quarter of the global DRAM market, Gartner forecasts that AI-related chips could represent more than half of total semiconductor revenue by 2029. 2026-01-14 17:09:29
  • Trumps Taiwan chip big deal reshapes foundry race, puts Korea in strategic bind
    Trump's Taiwan chip 'big deal' reshapes foundry race, puts Korea in strategic bind SEOUL, January 13 (AJP) - After extracting concessions from Japan and South Korea, U.S. President Donald Trump is now pressing another semiconductor powerhouse — Taiwan — using trade leverage to pull advanced chip manufacturing onto U.S. soil. Under an emerging deal, Washington plans to cut tariffs on Taiwanese goods to 15 percent from 20 percent in exchange for a major expansion of Taiwan Semiconductor Manufacturing Co.’s U.S. manufacturing footprint, according to a report by The New York Times. TSMC would commit to building at least five additional semiconductor facilities in the United States. The arrangement reflects Trump’s broader strategy of tying trade relief to domestic investment, particularly in industries deemed critical to national security. U.S. officials have indicated that companies expanding production in America could be exempted from potential national-security tariffs imposed under Section 232 of U.S. trade law. For TSMC, the deal would significantly accelerate its transformation of Arizona into a mega-cluster for advanced chips. The Taiwanese chipmaker is already operating one plant in the state and is completing a second scheduled to open in 2028, with more fabs planned for the late 2020s. The new commitment would lift TSMC’s total U.S. investment to an estimated $165 billion, covering advanced manufacturing, packaging facilities and research centers. Samsung Electronics, TSMC’s closest rival in contract chipmaking, has a smaller but expanding U.S. presence. The Korean tech giant operates two fabs in Austin, Texas, and is preparing to begin production at a new foundry in Taylor, Texas, where it has shifted its focus to next-generation two-nanometer technology. The Taylor site is designed to accommodate multiple fabs, though only one is currently nearing completion. The widening gap in U.S. capacity between the world’s two largest foundries comes at a sensitive moment for South Korea. The country’s semiconductor exports hit a record $173.4 billion last year, with shipments to the United States accounting for nearly one-fifth of the total amid surging demand for AI-server memory. Industry watchers say the Trump-TSMC deal could pull Korea’s chipmakers in opposing directions. On one hand, rising geopolitical and tariff risks surrounding Taiwan may prompt major U.S. customers to diversify supply chains, potentially benefiting Samsung’s foundry business as an alternative production base. On the other, TSMC’s push to build a fully integrated “all-in-U.S.” ecosystem — spanning manufacturing, advanced packaging and R&D — is expected to further lock in American customers, raising barriers for rivals seeking new contracts. The policy shift is also likely to intensify competition for investment dollars. If the U.S. manufacturing race accelerates, Korean chipmakers may need to significantly increase North American capital spending beyond current plans, adding pressure to cash flow at a time when the industry is already investing heavily in advanced packaging and high-bandwidth memory for artificial-intelligence systems. For now, officials in Seoul are closely watching how Washington formalizes the deal — particularly the scope of tariff exemptions under Section 232 and how U.S. authorities define a qualifying “facility.” Those details could determine whether Trump’s Taiwan chip “big deal” becomes a catalyst for broader supply-chain diversification — or a turning point that further entrenches TSMC’s dominance on American soil, forcing Korea’s chip champions to rethink their long-term strategy. Kim Yang-paeng, a senior researcher at the Korea Institute for Industrial Economics and Trade, said it was difficult to draw a direct line between Washington’s demands on Taiwan and their impact on South Korea, but warned of growing pressure on the global semiconductor ecosystem. “The fact that the U.S. is asking Taiwan to build more fabs does not necessarily mean the same demands will automatically apply to South Korea,” Kim said. “Korea already has broader industrial cooperation with the United States in areas such as shipbuilding and automobiles, which puts it in a different position from Taiwan.” Still, he cautioned that a rapid expansion of TSMC’s U.S. manufacturing capacity could exacerbate oversupply risks and accelerate the shift of global semiconductor production toward America. “If the U.S. increasingly becomes the center of chip manufacturing,” Kim said, “the room for Korea to expand its foundry business over the long term could narrow.” 2026-01-13 16:30:03
  • Strong memory prices, weak won fuel electronics inflation in Korea
    Strong memory prices, weak won fuel electronics inflation in Korea SEOUL, January 12 (AJP) - Soaring memory chip prices, compounded by a weak won, are pushing up the cost of electronics in South Korea, from laptops and smartphones to AI-driven devices such as robots and autonomous systems. All computing devices rely on memory chips, and the rapid expansion of hyperscale data centers and artificial intelligence has tightened global supply. As chipmakers prioritize high-performance memory for AI servers, device manufacturers and retailers are scrambling to secure components, driving up prices across the electronics market. “Simply put, today is the cheapest day to buy a laptop,” said a vendor at Seoul’s Yongsan electronics shopping mall, a longtime barometer of consumer electronics demand. Woo, a manager at Seoul’s Yongsan electronics shopping mall, said the shortage stems from chipmakers diverting conventional DRAM capacity to high-bandwidth memory used in AI workloads, leaving limited supply for consumer devices. “Demand is surging, but inventories are tight, and buyers are willing to pay a premium to secure components,” he said. “That has effectively doubled prices. This has been going on for the past two to three months, and I expect the market to remain tight at least through April or May.” Despite rising prices, sales have yet to slow meaningfully, he added. “Even with higher costs, people who need PCs right away are still buying,” Woo said. “At our store alone, around 100 custom PC orders were registered over the weekend, and in some cases deliveries are being delayed because supply cannot keep up.” Many consumers, however, are hesitating. Kim Beom-jin, an office worker browsing during his lunch break, said he recently postponed a purchase. “I was looking to buy a laptop, but when I came back, I was shocked by the price,” he said. “Now I’m trying to decide whether this is the right time to buy or if I should wait.” Memory prices have risen sharply since the second half of last year. The average selling price of DDR eight-gigabit DRAM jumped from about $1.3 in March to roughly $8 in November, as major producers — Samsung Electronics, SK hynix and Micron — shifted capacity toward high-bandwidth memory for hyperscale servers and AI accelerators. The increase in contract prices has already filtered through to device makers. PC manufacturers including HP and Dell have raised prices, citing higher component costs. Market researcher TrendForce said the impact is becoming increasingly visible. The firm forecasts that PC DRAM contract prices could rise by as much as 60 percent in the first quarter of 2026, a move expected to translate into 5 to 15 percent increases in retail prices for laptops and smartphones this year. Import costs are adding to the pressure. Statistics Korea data show prices for computer equipment and peripherals rose 15.2 percent in the fourth quarter of 2025 from the previous quarter, while semiconductor components, including memory chips, jumped 28.4 percent from a year earlier. The increase has been amplified by a weak won, as the currency hovered around 1,480 versus the U.S. dollar in December, losing more than 8 percent from 1,362 won in late June. Supply constraints remain tight. TrendForce estimates that PC DRAM prices rose 15 to 20 percent between the fourth quarter of 2025 and the first quarter of 2026, while server DRAM prices surged more than 25 percent amid booming demand from AI data centers. With chipmakers allocating more than forty percent of wafer capacity to high-bandwidth memory, supplies of conventional DRAM for PCs and smartphones have tightened. The impact is increasingly visible at retail. Average prices for gaming laptops equipped with Nvidia’s RTX 4060 have risen 25.5 percent since early 2024, while the cost of a standard 32-gigabyte DDR5 memory kit has jumped 62.5 percent over the same period. Mid-range custom PC builds are now about thirty percent more expensive than two years ago. Smartphone prices are also climbing. Samsung’s next flagship Galaxy S model, expected to launch in 2026, is projected to be priced at around 1.93 million won, about 13 percent higher than the 2024 version, according to industry estimates. Apple’s iPhone Pro models are expected to be priced nearly twenty percent higher than two years earlier. Kim Duk-gi, a professor of semiconductor systems engineering at Sejong University, said the current price surge reflects the inherently cyclical nature of the memory industry, now amplified by an unprecedented wave of AI investment. “Memory is produced on a continuous, twenty-four-hour cycle, unlike logic chips that are designed to meet specific demand,” Kim said. “With the rapid expansion of AI data centers, demand has surged much faster than supply can adjust. That is why we are now in a shortage phase, but structurally, the industry will eventually swing back to oversupply.” Kim added that the tight market is unlikely to ease quickly. “The shortage is still relatively recent, so it will be difficult to resolve in the near term,” he said. “For now, elevated prices are likely to persist.” Counterpoint Research estimates that in 2026, memory components could account for up to 35 percent of a device’s retail price, marking what it describes as the peak of “component-driven inflation” in the global electronics sector. 2026-01-12 16:58:33
  • Musk says Teslas humanoid robot could outperform top surgeons within three years
    Musk says Tesla's humanoid robot could outperform top surgeons within three years SEOUL, January 10 (AJP) - Tesla CEO Elon Musk said humanoid robots could surpass the world’s best human surgeons within three years, arguing that rapid advances in artificial intelligence and robotics will transform healthcare faster than many expect. Musk made the remarks on the “Moonshots” podcast hosted by U.S. physician and engineer Peter Diamandis, outlining his vision for Tesla’s Optimus robot and the future of medical care. “Right now there’s a shortage of doctors and great surgeons,” Musk said. “It takes a super long time to learn how to be a good doctor, and even then, the knowledge is constantly evolving. Doctors have limited time, they make mistakes.” Asked when Optimus could outperform the best surgeons, Musk replied: “Three years. And by the way, that’s three years at scale. There will probably be more Optimus robots that are great surgeons than there are all surgeons on Earth.” The SpaceX founder said the pace of development in humanoid robots should be viewed through what he called “exponential growth” in three areas — AI software capability, AI chip performance and mechanical precision — adding that Optimus could eventually become capable of performing complex procedures more consistently than humans. Some medical experts, however, have questioned Musk’s timeline. Arthur Caplan, a bioethicist at New York University’s Grossman School of Medicine, told The Independent that the claim that robots could outperform the best human surgeons within three years was “not credible,” citing the complexity of surgery and the difficulty of programming for human variability. Musk’s comments have also sparked debate on social media, where users have clashed over the implications of his remarks. While some expressed concern about the future of medical professions and patient safety, others questioned the credibility of Musk’s predictions, arguing that expectations for humanoid robots are running ahead of current technological reality. The Tesla CEO has previously said Optimus could enter commercial production in 2026, positioning humanoid robots as a central pillar of the company’s long-term strategy alongside electric vehicles and artificial intelligence. 2026-01-10 17:28:26
  • U.S. dietary guidelines recommend kimchi, overhaul federal nutrition advice
    U.S. dietary guidelines recommend kimchi, overhaul federal nutrition advice SEOUL, January 10 (AJP) - The United States government on Friday released updated 2025–2030 Dietary Guidelines for Americans that, for the first time, include fermented foods such as kimchi as part of recommended diets and mark a significant shift in federal nutrition policy. Health and Human Services Secretary Robert F. Kennedy Jr. and Agriculture Secretary Brooke Rollins jointly unveiled the guidelines at a White House briefing in Washington, saying the recommendations will guide federal nutrition standards over the next five years, including school meals, military food programs and assistance schemes such as SNAP (Supplemental Nutrition Assistance Program). The new guidelines emphasize eating whole, minimally processed foods and increasing intake of protein, healthy fats, vegetables, fruits and whole grains. They move away from diets centered on ultra-processed foods and added sugars, which are strongly discouraged, and raise the recommended daily protein intake from 0.8 grams to 1.2–1.6 grams per kilogram of body weight. A notable addition in the guidelines is the inclusion of fermented foods — including kimchi, sauerkraut, kefir and miso — under recommendations for supporting gut health alongside vegetables and high-fiber foods. The document cites the importance of diverse microbiome health in its dietary guidance. The guidelines also ease past negative assessments of animal-based foods such as red meat, eggs and seafood, and allow consumption of whole-fat dairy products and animal fats, while continuing to caution against dairy products with added sugars. Kennedy described the overhaul as a “return to real food” and part of the government’s broader effort to address chronic disease, which officials say is driven in large part by highly processed diets. The Dietary Guidelines for Americans are updated every five years by the Department of Health and Human Services and the Department of Agriculture. The 2025–2030 edition reflects what federal officials described as the most significant reset of national nutrition policy in decades. 2026-01-10 16:27:49
  • North Korea warns South Korea over alleged drone incursions
    North Korea warns South Korea over alleged drone incursions SEOUL, January 10 (AJP) - North Korea on Saturday accused South Korea of sending drones into its airspace last year and earlier this month and warned that Seoul would “pay the price” for what it called repeated violations of its sovereignty. In a statement carried by the Korean Central News Agency, a spokesperson for the North’s military general staff said a South Korean drone was detected on Sunday flying northward over the border area near Ganghwa Island before being shot down and forced to crash near Kaesong. The spokesperson also claimed that another drone launched from Paju in Gyeonggi Province on Sept. 27 last year crossed into North Korean airspace over Hwanghae Province and was brought down through electronic warfare. North Korea said the downed drones carried surveillance equipment and video footage of areas inside the country. South Korea’s defense ministry rejected the claims, saying its military did not operate drones on the dates cited by Pyongyang. “Our military did not conduct any drone operations on the days North Korea claims,” the ministry said on Saturday, adding that the drones shown by North Korea were not models operated by South Korea’s armed forces. Defense Minister Ahn Gyu-baek said the allegations were “completely untrue” and suggested that the two Koreas could conduct a joint investigation if necessary. The presidential office said it convened a National Security Council working-level meeting to assess the situation, adding that President Lee Jae-myung had ordered a thorough investigation into the matter. North Korea said drone activities by South Korea near the border had continued even after a change of government in Seoul, calling the South its “most hostile enemy” and warning that further actions would not be tolerated. 2026-01-10 14:37:59