Japan is stepping up its push into physical AI: at CEATEC 2026, major technology companies and research institutions showcased advances in autonomous robot learning, motion sensing, and agricultural automation, with Murata, AWS, and Hitachi presenting self-learning robots. The source does not state the cause and contains no price, capacity, or supply-demand data.
physical AIself-learning robotsmotion sensingMurata, AWS and 3 totalDigiTimes Asia
TSMC reported record earnings for the third quarter of 2026 and forecast a further 15% increase in fourth-quarter revenue, putting full-year US dollar revenue growth on course to reach 44%. Supply chain sources say the earnings call carried a more consequential message: TSMC's increasingly explicit assertion of the bargaining power afforded by its lead in advanced process technology and manufacturing scale.
SK hynix presented its next-generation AI infrastructure vision at the OCP Global Summit 2026 on October 16, 2026, stating that "Memory-Centric Server Systems Are the Future". The body text lists only related keywords, including Agentic AI, AI, CXL, DRAM, eSSD, HBF, HBM and OCP Global Summit, and provides no further details such as specific product specifications, performance figures or commercial arrangements; the source does not state the cause of any related technology or market developments.
SK hynix took part in OCP Global Summit 2026, held from the 12th to the 15th of October 2026 in San Jose, California, where it unveiled a wide range of next-generation memory solutions and presented the direction memory technology should take in the AI era. At the exhibition, SK hynix unveiled its second-generation CMM-DDR5 (256GB, based on CXL 3.2), which per the source more than doubles bandwidth (80GB/s) and improves energy efficiency by about 30% compared with the previous generation, and demonstrated CXL pooled memory in AI agent serving and KV cache-sharing scenarios. The source states the background cause: as the volume of data handled by AI grows explosively, data bottlenecks have emerged as a new challenge. SK hynix also showed its product portfolio spanning HBM4/HBM4E/HBM3, HBF, server DDR5 RDIMMs (256GB/128GB/96GB/64GB) and eSSDs. The article also mentions that in September last year SK hynix signed a business agreement with NAVER Cloud, and that the PS1110 E3.S and PS1101 E3.S eSSDs are currently under sequential development.
CXL Pooled MemoryCMM-DDR5MARSSK hynixSK Hynix Newsroom
At the OCP Global Summit 2026, Nvidia showcased the latest developments in its MGX server platform ecosystem, unveiling for the first time a switch board sample using polytetrafluoroethylene (PTFE) and naming seven supply-chain partners. Taiwan's Flexium Interconnect and Taiflex Scientific joined Chinese suppliers Victory Giant Technology, Shennan Circuits, Shenzhen Kinwong Electronic, Shengyi Technology, and US-based TTM Technologies on the partner list. The source does not state the cause.
PTFE switch board sampleNvidia, Flexium Interconnect and 8 totalDigiTimes AsiaPartnership/M&A
At a SEMICON West 2026 industry outlook session, Duncan Meldrum, an economist at Hilltop Economics, warned that while investment in AI data centers remains strong, semiconductor growth is becoming increasingly concentrated in cloud AI, and the long-running decline in consumer technology prices has come to an abrupt halt; he said this shift is putting additional pressure on the broader semiconductor market.
non-AI semiconductorscloud AI semiconductorsDigiTimes Asia
Taiwan Semiconductor Manufacturing (TSMC) is seeing rising demand for CPUs related to agentic artificial intelligence (AI), but has not yet included CPUs in its calculations of AI accelerator demand growth. TSMC chairman C.C. Wei said at the company's earnings conference on October 15 that the company is maintaining its original, narrower definition of AI demand, mainly because it is difficult to clearly distinguish whether CPUs are used in traditional servers or AI servers. Actual demand also depends on how customers deploy the technology and configure their products.
Chinese semiconductor technology start-up Langdao Silex (Shanghai) Technology has publicly unveiled an extreme ultraviolet (EUV) light source prototype with a claimed 90% localization rate and 15W output power. The source does not state the cause. The company frames this as progress in China's efforts to develop domestic lithography technology; however, the system remains three to five years from mass production, with output power and long-term stability posing major engineering challenges. The event is a prototype debut and does not represent mass production or commercial use.
EUV light sourceLangdao Silex (Shanghai) TechnologyDigiTimes Asia
LG Innotek expects its semiconductor substrate business to improve in 2027. The company is pursuing long-term supply agreements (LTAs) with major technology companies and preparing to enter the high-end AI accelerator market. Under this strategy, it plans to secure customer commitments before expanding production capacity as it seeks to diversify beyond smartphone camera modules. These items remain plans/expectations and do not constitute an implemented capacity expansion.
semiconductor substratesmartphone camera modulesLG InnotekDigiTimes AsiaCapacity Change
Samsung Austin Semiconductor (SAS) reported that in 2025 its Austin and Taylor campuses injected a total of $10.9 billion into the Central Texas economy, supporting 28,746 jobs and paying $2 billion in wages. The campuses generated $218.2 million in annual tax revenue for local taxing districts, with vendor payments of $1.9 billion at the Austin campus and $7.4 million at the Taylor campus. Charitable contributions totaled nearly $4 million and workforce development investments amounted to $5.9 million. In 2026 SAS aims to establish the Taylor fab using extreme‑ultraviolet (EUV) technology to produce 2 nm chips in the U.S., expecting the site to be operational by the end of 2026, and plans to invest more than $37 billion in the region in the coming years.
According to Wccftech, a South Korean semiconductor engineer using the X handle @TTHUEONE accused Tesla of demanding access to Samsung's internal fab information and audit rights for its AI5 and AI6 chip contracts while aggressively pushing down foundry prices; the engineer also alleged that Terafab had poached numerous employees from Samsung's Taylor fab in Texas and that Tesla turned to Samsung Foundry for its AI5 and AI6 chips because TSMC would not meet its demands. Earlier this week, Musk praised Micron, saying the company 'has been reasonable regarding future memory prices, while others have been extortionate,' remarks widely interpreted as criticism of Samsung. As background: Business Post warned in March that Terafab (planned site about a 30-minute drive from Samsung's Taylor foundry) could intensify competition for talent and equipment and flagged risks to Samsung's intellectual property and manufacturing know-how; Financial News reported on October 12 that Samsung has begun deploying advanced-node yield engineers from South Korea to its Taylor facility while expanding local recruitment ahead of planned 2nm mass production in 2027, aiming to begin operating the fab by the end of 2026, with Tesla among key customers; Seoul Economic Daily reported in September that Samsung had begun trial production of Tesla's AI5 chips at the facility with yield verification underway ahead of full-scale production in 2027. AsiaToday earlier reported that Tesla's AI5 and AI6 orders are expected to support initial capacity utilization at the Taylor fab, and that TSMC's potential involvement in Terafab could complicate Samsung's efforts to secure future chip orders. TSMC Chairman C.C. Wei confirmed the company is in discussions over Terafab, though the outcome remains uncertain, and emphasized that any fab involving TSMC would have to be operated by the company itself; per Reuters, Musk said on October 7 that his companies would build and operate Terafab, with TSMC potentially subleasing part of the facility rather than taking an operational role.
AI5 chipAI6 chip2nm processTesla, Samsung and 6 totalTrendForcePartnership/M&A
According to Reuters, SK hynix is reportedly asking the U.S. government to authorize additional semiconductor manufacturing equipment shipments to its Dalian facility in China in exchange for a commitment to build a new NAND fab in the U.S. The Dalian fab is the world's only facility to produce 192-layer floating gate 3D NAND for Solidigm's data center-grade SSDs for AI, enterprise, and hyperscale data centers; its manufacturing operations rely on equipment from American suppliers including Applied Materials, KLA, and Lam Research, whose exports to China require a U.S. government license. Solidigm has requested U.S. export licenses for additional tools, but the U.S. government has yet to authorize the shipments; last December the U.S. Commerce Department granted SK hynix a limited authorization for equipment deliveries to Dalian for 2026, but has not issued a similar equipment export authorization for 2027. Solidigm wants to install additional manufacturing equipment in the unused portions of its Dalian Mod B facility, completed in 2023, and is also considering an initial public offering as early as 2027 to raise additional capital. The proposed U.S. fab and the equipment export arrangement remain reported proposals, not approved or implemented actions.
3D NAND192-layer floating gate 3D NANDdata center-grade SSDsSK hynix, Solidigm and 5 totalTom's HardwarePolicy/Regulation
Related manufacturers
SK hynix, Solidigm, Applied Materials, KLA, Lam Research
According to ZDNet, SK hynix has unveiled its ultra-lightweight Hynix Transaction Processing Unit (H-TPU) architecture, dedicated to SSD data-plane control and transactions, distributing storage-control tasks across dozens of compact H-TPU clusters. According to Vice President Koo In-jae, who oversees SoC design, H-TPU eliminates complex CPU features such as superscalar and out-of-order execution and adopts a simplified four-stage pipeline, reducing processor area to approximately 4% and power consumption to around 2.5% of a conventional CPU core, activating only required modules. Building on H-TPU, SK hynix plans to introduce hardware automation (HWA), assigning repetitive storage-control tasks such as instruction processing, data address lookups, and temporary buffer copying to dedicated hardware circuits. SK hynix expects the H-TPU-based controller architecture to complement high-bandwidth flash (HBF), an emerging memory technology for AI workloads; AI inference is highly read-intensive with read-to-write ratios reaching approximately 300,000 to 1, and AI compilers and runtimes could inform the HBF SoC of upcoming data-processing needs for prefetching. Separately, according to Yonhap News, Samsung began mass production of its PCIe 6.0-based PM1763 enterprise SSD in July 2026, featuring a newly developed 4nm controller and ninth-generation V-NAND, delivering more than twice the performance and over 1.8 times the power efficiency of its predecessor.
Samsung Electronics acknowledged at the 2026 JEDEC Forum, held alongside SEDEX 2026 at COEX in Seoul on Oct. 15, that standardizing next-generation AI memory faces challenges: customer-specific designs extend beyond the base dies of high-bandwidth memory (HBM) into DRAM core dies, making them increasingly difficult for memory manufacturers to manage. Hwang Ga-ram, a researcher on the product planning team of Samsung's Memory Business under its DS division, said LPDDR-PIM and zHBM have many customer-specific features that are extremely difficult to manage in actual production, and added that other memory manufacturers likely face the same challenges. Samsung's proposed zHBM architecture uses a 3D stacked structure that directly connects HBM to the computing processor, with TSVs distributed throughout DRAM core dies, so customized designs extend beyond the base die to the overall structure, presenting even greater standardization challenges than conventional 2.5D packaging. Hwang expressed hope that the LPDDR6-PIM standard would be established by the end of this year, and said not every function can be fully standardized while hoping JEDEC member companies could cooperate on common standards. Samsung also disclosed LPDDR5X-PIM development results: memory bandwidth is eight times that of conventional LPDDR5X; AI inference performance reached 102 TPS, approximately 3.8 times the 27 TPS of LPDDR5X, an improvement of about 23% over the 83 TPS demonstrated at Hot Chips in August; average power consumption was 11.5 watts, approximately 36% higher than the 8.4 watts of LPDDR5X; and inference performance per watt improved to 9 TPS/W, roughly 2.8 times the 3.2 TPS/W of conventional LPDDR5X.
Iron Device announced on Oct. 15 that it will join a government-backed research project led by South Korea's Ministry of Trade, Industry and Resources (MOTIR) to develop an on-device AI computing module for robotic hands. The project has a total budget of approximately 5.25 billion won and will run from August this year through the end of 2029. Iron Device will develop a low-power system-on-chip (SoC) integrating CPU, GPU and NPU, aiming to reduce the module's footprint by about 59% compared with Nvidia's Jetson Orin Nano while tripling power efficiency (TOPS/W) and cutting latency from sensor data to motor control to 20 ms or less. The company plans to sell the completed module to domestic and overseas humanoid and industrial robot customers independently of the project, targeting a reduction of customers' robot system costs by more than 25%.
XMax Inc of Commerce, CA, USA (formerly Nova LifeStyle Inc), which designs, sources and distributes contemporary furniture but has diversified to also developing and providing artificial intelligence software and platform-based services, has agreed to acquire Hexa Creation Inc. Hexa Creation is a US-based company focused on 1200V vertical gallium nitride (GaN) power semiconductor technology for AI data-center power infrastructure and other high-voltage applications. The agreement has been reached, but the source does not state a completion timeline or whether regulatory approvals are required.
VON ARDENNE announced that it is expanding its semiconductor deposition portfolio into panel-level packaging (PLP) with a new HiPIMS deposition solution. The solution deposits high-aspect-ratio barrier and seed layers in through-glass via (TGV) holes, enabling dense, highly adherent functional film deposition, and leverages the company's HISS modular horizontal coating platform, scaling HiPIMS from 300mm wafer-level processing to panel-level production for advanced packaging applications including TGVs and redistribution layers (RDL). According to Yole Group, PLP generated more than $300M in revenue in 2025 and is projected to grow by more than 40% CAGR over the next five years; per a March 2026 Yole Report titled 'Panel Level Packaging 2026,' PLP is set for strong acceleration toward 2029-2030, driven by TSMC's entry and adoption of high-end PLP for AI and high-performance computing (HPC) applications. VON ARDENNE executives said HiPIMS can reduce downstream costs and enhance yield. This is a product/portfolio expansion announcement; the source does not state production timing or implementation details.
panel-level packaging (PLP) deposition equipmentthrough-glass via (TGV) barrier and seed layersredistribution layers (RDL)VON ARDENNEEE Journal
According to an Oct. 14 Bloomberg report cited by the article published 15 October 2026, the EU is in talks with China to lift restrictions China imposed on Sindlhauser Materials GmbH, a German precision-manufacturing and advanced-coatings firm in ASML's supply chain; in return, the EU would remove restrictions it placed on one unnamed Chinese firm accused of aiding Russia. The swap was discussed during EU Commissioner for Trade and Economic Security Maroš Šefčovič's visit to Beijing; the EU is expected to make the first move, with any delisting on its end requiring unanimous support from member states, and both sides agreed to share more information and give advance warning on future export restrictions. In July 2026, China put Sindlhauser, along with 13 other European companies, on its export-control list in apparent retaliation for the EU's 21st Russia sanctions package, which sanctioned several Chinese firms accused of supplying Moscow with components usable for weapons; Beijing's official reason was the firms' alleged dealings with Taiwan. The restrictions ban Chinese exporters from selling dual-use items to the firm and prohibit foreign companies and individuals from passing on China-origin dual-use items to it, blocking third-party sourcing. Sindlhauser relies almost entirely on Chinese rare earths and critical minerals; the German and Bavarian state governments have helped explore workarounds such as alternative suppliers and recycling existing materials, but the company risks going out of business in the coming months if the sanctions remain.
Chinese rare earthscritical mineralsdual-use itemsSindlhauser Materials GmbH, ASMLTom's HardwarePolicy/Regulation
Siemens announced significant advances in its collaboration with NVIDIA for computational lithography: by integrating NVIDIA cuLitho, a CUDA-X library for GPU-accelerated computational lithography, with Calibre manufacturing software, the collaboration delivers up to 12x acceleration for full-chip optical proximity correction (OPC) workloads and at least 10x acceleration for key lithography simulation workloads; verified engineering data supports simulation speed-ups ranging from at least 10x to 17x for key workloads. It enables practical full-chip optimization of photomasks for High-NA EUV lithography using advanced methods such as Inverse Lithography Technology (ILT), addressing the critical runtime and complexity challenges of leading-edge patterning for AI compute and memory devices. Siemens says this can help reduce simulation runtimes, identify lithographic hotspots earlier, improve process margins before tapeout, and support the deployment of new AI silicon. The source does not state any impact on specific prices or supply-demand conditions.
In a Power Corner interview, Jason Ngai, senior product marketing manager at pSemi, discussed the ARC1C0608, an LED backlight driver suited for single-cell (1S) Li-ion laptops that need a high conversion ratio from 2.7 to 5.5 V input to LCD backlight voltage of up to 30 V. According to Ngai, as modern SoCs have become highly efficient in standby, the display has become the dominant power draw, consuming up to 70% of total system power, and it cannot be put into standby—only dimmed or shut off. He explained that LED backlight drivers are boost converters whose efficiency worsens as input voltage falls: a typical 1S LED driver runs around the mid-to-high-80 percent range but drops as low as the mid-60s to mid-70s percent across the battery range (about 4.35 V fully charged down to about 3.2–3.6 V near cutoff). The ARC1C0608 uses a proprietary two-stage architecture combining a boost stage and a charge pump, achieving 95% peak efficiency and high-80s to over 90% efficiency across most of the operating range; in a 3.8-V input, 22–24 V output example, a 3x charge pump lets the boost stage reach only about 8 V. This reduces reliance on the inductor, enabling a low-profile solution with a 1.2-mm chip-size inductor (or 0.6-mm height with only about 1% efficiency loss). Beyond laptops, the IC is also suited for other single-cell display applications such as tablets, VR headsets, drone headlights, and bicycle displays.
LED backlight driverLCD backlightpSemiPower Electronics News