AX·NEWS
Sign in

Semiconductors

Semiconductors news — Tuesday 27 October 2026

150 stories from 12 sources, every one in English. Each links to its publisher.

Open in the live feedRSS feedOn a TV

  1. OpenAI, a Longtime Nvidia Customer, Plans to Buy Google’s AI Chips
    USThe Information

    OpenAI plans to buy Google ’s AI chips, according to a person familiar with the matter, in a boost for a rising competitor to OpenAI backer Nvidia. Financial terms and other details of OpenAI’s planned purchase couldn’t be learned. But any…

  2. Is Anduril Worth $120 Billion?
    USThe Information

    Some senior employees at Anduril Industries have indicated to investors that the autonomous weapons maker hopes to raise a new funding round in the first quarter next year, I’ve learned. It has considered doing so at a valuation of around…

  3. Transitioning Legacy Aircraft Power Architectures to High-Density Switching Supplies
    USEE Times

    Aircraft electrical systems often stay in service for decades, even as avionics demand more from established interfaces. High-density switching supplies offer one way to modernize legacy aircraft power

  4. With China having fewer kids, dairies rethink milk for adults
    JPNikkei Asia
  5. Statesman: Stephen Schwarzman (27)
    JPNikkei Asia
  6. Microsoft Maia 300 Shown at OCP 2026
    USServeTheHome

    At OCP Summit 2026, we saw the new Microsoft Maia 300 platform with four 288GB HBM4 950W accelerators per tray The post Microsoft Maia 300 Shown at OCP 2026 appeared first on ServeTheHome .

  7. Japanese education group Gakken returns to India with accent on values
    JPNikkei Asia
  8. US to halt visa waivers for Chinese traveling to Pacific territory
    JPNikkei Asia
  9. As AI revives magnetic tape, Fujifilm plans biggest buildup in two decades
    JPNikkei Asia
  10. India's rise past Japan is a wake-up call for both nations
    JPNikkei Asia
  11. Analog Superposition Increases Effective PCM Storage Density (IBM Research Europe)
    USSemiconductor Engineering

    Researchers at IBM Research Europe published “Superposed analog memory for vector storage in neural networks.” Summary : Researchers demonstrated an analog phase-change memory architecture that stores multiple AI vectors in superposition…

  12. Researchers develop wearable sensors to track movement during sports and exercise
    GBTechXplore Semiconductors

    What does a good biceps curl look like? Walk into a busy gym, and the answer may vary from one person to the next. Some people learn by watching a trainer, while others copy fellow gym-goers or follow videos on their phones. Without…

  13. Fields Medal-winning UC Berkeley professor ends graded homework due to AI cheating
    USTom's Hardware

    A University of California, Berkeley (UC Berkeley) professor has stopped giving his students graded homework, saying that rampant AI use has made it pointless. According to the New York Post , renowned mathematician Richard Borcherds said…

  14. Takeda debuts narcolepsy drug targeting root cause, capping 15-year effort
    JPNikkei Asia
  15. HP and Asus shift some PC making back to China due to AI-driven cost hikes
    JPNikkei Asia
  16. OpenAI, Anthropic Claim Cryptography Breakthrough on Same Day
    USThe Information

    OpenAI’s models have disproved a foundational theory within cryptography, the AI company claimed on Monday. OpenAI also announced two other breakthroughs in physics and mathematics. The findings come after OpenAI released over 700 math…

  17. Flash heating offers faster way to recover critical metal from e-waste
    GBTechXplore Semiconductors

    Many electronic devices use flame-retardant plastic as casing in order to reduce the risk of electrical fires. This plastic is formulated using a metal called antimony along with organic bromide compounds. In fact, 39% of the world's…

  18. Samsung-linked hospital to bring humanoid robots into operating room
    JPNikkei Asia
  19. Japan banks, fintech step up digital currency push for trade settlements
    JPNikkei Asia
  20. BYD rolls out public test tracks to promote China's car culture
    JPNikkei Asia
  21. Why Cheaper Models Won't Win OpenAI the Enterprise
    USThe Information

    OpenAI is trying to chip away at Anthropic’s lead with enterprises by fashioning itself as a cheaper, more flexible provider, Steph and I reported this weekend . But Anthropic's impressive prospectus for an initial public offering , filed…

  22. Oracle To Start Running Google TPUs in Its Cloud
    USThe Information

    Google will sell its tensor processing units to Oracle and OpenAI, challenging Nvidia’s hold on the AI chip market. OpenAI is planning to buy Google’s TPUs, The Information reported . Google on Tuesday also said it would partner with…

  23. Military drone outfit Anduril partners with Qualcomm to build laptops on US soil with new BlueChip venture
    USTom's Hardware

    Defense contractor Anduril, which is known for its military drones, just announced that it’s building laptops in the U.S., with the company saying that it will begin taking orders in 2027. According to its press release , it will call its…

  24. Israeli ad tech firm tied to suspected Starlink tracking effort, report claims
    USTom's Hardware

    Discover-Tech, an Israeli advertising technology company, has reportedly been linked to a suspected surveillance effort aimed at Starlink users, according to Reuters . The report, which cites six people familiar with the matter, claims the…

  25. North Korea places self-propelled guns near DMZ
    JPNikkei Asia
  26. Japan's EV subsidies face uncertain future as country's 'DOGE' takes aim
    JPNikkei Asia
  27. Japan's Sumitomo Electric offers more low-tungsten tools to cut China reliance
    JPNikkei Asia
  28. Chinese media say Japan TV show 'Vivant' torture scenes 'stoke hostility'
    JPNikkei Asia
  29. Your Robot’s Safety Functions Already Work. What If the Input Lies?
    USIEEE Spectrum

    This article is brought to you by VicOne . A mobile robot slows for someone in a hallway. A collaborative arm eases its movement as a worker approaches. A humanoid pauses to let a person pass. Each response depends on information about the…

  30. Why Synopsys’ Role in AMD’s Instinct MI455X Matters for AI Infrastructure
    USSemiWiki

    Synopsys’ collaboration with AMD on the Instinct MI455X GPU highlights a shift in how advanced AI processors are developed. Increasing performance depends on coordinating compute dies, memory interfaces, packaging, power delivery, and…

  31. TSMC reportedly weighs 20% CoWoS price increase as Nvidia Blackwell and Rubin chips put pressure on packaging capacity
    USTom's Hardware

    TSMC is weighing raising prices for its Chip on Wafer on Substrate (CoWoS) advanced packaging technology by as much as 20%, reports Trendforce , citing supply chain sources and local media. The increase reflects strained capacity amid AI…

  32. Why Anthropic's 85% Margin Is a Reality Check for Skeptics
    USThe Information

    When Anthropic on Monday dropped its long-awaited initial public offering prospectus filing, a lot of the details—see our top takeaways here —weren’t exactly new, given how much my colleagues and others had previously reported about it…

  33. Samsung Portable SSD P9 review: speedy USB4 in a slim, pricey package
    USTom's Hardware

    It’s a tough time to launch an external SSD, with demand for flash and RAM at all-time highs thanks to AI hyperscaler demand. But content creators, videographers, and media collectors need more, faster, and roomier options, so Samsung has…

  34. CopperPilot Raises Pre-seed to Develop AI Agent for PCB Design
    USEE Times

    Hardware design startup CopperPilot has raised $390,000 in a pre-seed funding round led by Neon Fund, with participation from U.S.-based angel investors, as it opens version 1 of its AI design agent,

  35. Police are reportedly pulling motorists over because AI algorithms deem their driving patterns 'suspicious'
    USTom's Hardware

    Police are increasingly using AI-derived pretext to justify pulling drivers over, claims a report by the American Civil Liberties Union (ACLU) on October 26. According to Jay Stanley, a senior policy analyst with the ACLU Speech, Privacy…

  36. Walmart slashes $750 off this MSI gaming laptop — 16-inch Crosshair sports an RTX 5060 GPU and a fast 240Hz QHD display
    USTom's Hardware

    Spotted at Walmart today, at a pretty good price, is this RTX 5060-powered MSI Crosshair gaming laptop for just $1,249 . With price rises across the board on laptops and PCs, it's important to keep an eye out for large discounts on gaming…

  37. Indonesia-Singapore spats escalate over haze, energy, MSCI
    JPNikkei Asia
  38. Meta Slightly Increases Equity Fundraising Plan to $97.5 Billion
    USThe Information

    Meta Platforms is slightly increasing the size of its planned stock sales to fund AI investments to a total of $97.5 billion from $95 billion, the company said. The company is increasing the common stock offering in its fundraising plan to…

  39. Keybytes ‘truly modular keyboard’ starts with one key and is assembled like Legos
    USTom's Hardware

    Developer notaroomba (AKA Nathan Alspaugh) has published details of his Keybytes project on GitHub . In essence, this is a modular project where a Keyword hub is magnetically connected to Keybit keys, and each key has its own…

  40. Apple adds M6 to 14-inch MacBook Pro and iMac
    USTom's Hardware

    Apple is updating two more computers to its M6 system-on-chip. The base 14-inch MacBook Pro and the iMac lineup are getting the company's latest processor ahead of tomorrow's "Pro and Tell" event in Los Angeles. These devices are retaining…

  41. Hacker awarded $50k bounty for reporting bug to Sony that allowed PS5 Linux exploit
    USTom's Hardware

    A hacker used an AI LLM to find a hypervisor bug that allowed security researchers to load Linux on PlayStation 5 consoles , subsequently reported it to Sony, and received a $50,000 bug bounty. IGN reports that the person who discovered…

  42. Can AI Help Ray Kurzweil Cheat Death?
    USThe Information

    Ray Kurzweil does not want to die, but at the age of 78, he has been thinking more about it, a preoccupation that he describes as a “major motivating force of human beings.” So over the past year, Kurzweil, whose keen observations and 2005…

  43. S2C Targets Faster RISC-V Bring-Up With Prototyping and AI
    USSemiWiki

    S2C showcased FPGA prototyping and AI-assisted debugging at the sixth RISC-V Summit China, outlining how developers can validate complex processor designs and begin software development before first silicon. Its demonstrations connected…

  44. I Used AI to Fake a Scientific Paper. Here’s What Happened
    USIEEE Spectrum

    Science has a spam problem. And the spam filter is getting clogged up. According to all reports, academic publishing is drowning in what has come to be called AI slop: low-quality, mass-produced content generated with AI. By 2025, an…

  45. Open3DLab accuses Roblox of downloading 145GB of 3D assets for AI training purposes
    USTom's Hardware

    Independent 3D modeling community Open3DLab has publicly warned that a web crawler, linked to the gaming platform Roblox, mapped "almost every project" on its website, along with sister sites SFMLab and SmutBase, downloading 145.5GB of…

  46. Pop Mart's revenue halves in Q3 as Labubu craze cools
    JPNikkei Asia
  47. Intel's APX offers 'better performance' in Windows 11 according to new Insider setting
    USTom's Hardware

    A new "Advanced Performance Extensions (APX)" option has been spotted in a version of Windows 11 made available to Windows Insiders on October 23, with the feature set to enable a boost in application performance on compatible chips. At…

  48. Grab 32GB of DDR5 RAM for just $151 in this 4-in-1 Newegg bundle with the 9800X3D and a 2TB SSD
    USTom's Hardware

    Rather than spending $450 on 32GB of DDR5 all by itself, you can save a ton of money when you pick up RAM as part of Newegg's PC component bundles, $350 in this case. In this bundle, you can pick up four parts you need for $1,180.99 in a…

  49. Nvidia is bankrolling AI accelerator supply chain buildout to reserve production capacity
    USTom's Hardware

    On October 21, during GTC Berlin, Nvidia announced several massive, multi-billion-dollar deals with supply chain partners across substrates, advanced packaging, semiconductor testing, and optical engines to expand production capacity…

  50. 60,000 Microsoft employees offered voluntary buyouts as AI buildout costs bite
    USTom's Hardware

    Microsoft has offered voluntary buyouts to around 60,000 employees across the United States, Canada, India, and the United Kingdom. The employees, who must have an experience level of 70 or below, make up around 27% of the company's total…

  51. Anthropic prepares for potential $2 trillion IPO valuation after reporting billions in revenue
    USTom's Hardware

    Claude maker Anthropic has filed for an initial public offering (IPO) that could value the AI company at around $2 trillion, setting the stage for one of the biggest potential stock market debuts in the technology industry, the company…

  52. Data center gear is being hijacked in transit to the tune of hundreds of millions
    USTom's Hardware

    Have you been watching many heist movies lately? Here's a Hollywood-ready script for a new one: Organized crime teams have taken to stealing data center gear off or with their transport trucks, to the tune of tens of millions per shipment…

  53. Jensen and Musk both think aging GPUs are worth more than the books show
    TWDigiTimes Daily

    Speaking at GTC Berlin and on SpaceX's third-quarter 2026 earnings call, respectively, Nvidia CEO Jensen Huang and Tesla CEO Elon Musk reached a rare consensus on a pivotal topic: GPUs are no longer rapidly depreciating consumables…

  54. Tesla sees Cybercab becoming a new growth bet amid declining auto revenue
    TWDigiTimes Daily

    Tesla's automotive business weakened in the third quarter of 2026 as lower average vehicle prices and battery shortages pressured revenue and profit, while the company accelerated investment in artificial intelligence and its autonomous…

  55. 5 things to know about the US midterms and what they mean for Asia
    JPNikkei Asia
  56. Australia's AI build-out hits the grid: how a delayed power line wiped out AUD3 billion
    TWDigiTimes Daily

    Australia's push to become an AI infrastructure hub for the Asia-Pacific is running into constraints that capital alone cannot solve: the last mile of grid connection, and the consent of the communities that power lines run through. That…

  57. SpaceX closes $40 billion debt financing for GPUs expected in the first half of 2027, Musk says
    USTom's Hardware

    Elon Musk has confirmed that SpaceX closed $40 billion in debt financing for graphics processing units (GPUs), with the chips expected to arrive in the first half of 2027. He disclosed the deal on X, October 25 (embedded below) while…

  58. Quantum Computing’s Energy Bill Is Coming, and We Haven’t Priced It In
    USEE Times

    UN researchers project that by 2030 , data centers will use twice the power and water they used in 2025, largely because of AI. Energy has become a key constraint to consider in scaling up compute. The same

  59. Chinese EVs push imported models ahead of South Korean vehicles
    TWDigiTimes Daily

    Imported electric vehicles captured more than half of South Korea's new electric passenger-car registrations during the first nine months of 2026, marking the first time the share of domestically made models fell below that threshold…

  60. Pat Gelsinger warns AI compute efficiency must jump 10,000x amid power, memory, and fab constraints
    TWDigiTimes Daily

    Silicon Valley venture capital firm Playground Global has returned to Taiwan alongside a cohort of portfolio companies specializing in memory, power management, optical communications, and advanced packaging to showcase joint technical…

  61. Nvidia DSX POD challenges traditional AI data center construction
    TWDigiTimes Daily

    At its GTC conference in Berlin, Nvidia discussed the RTX Spark notebook, various implementations and partnerships involving physical AI, AI agents, and CUDA-X, as well as the DSX POD and European sovereign AI. Following its recently…

  62. Nvidia Invests $3 Billion in Meta in Deeper AI Partnership
    USThe Information

    Nvidia will invest $3 billion in Meta Platforms’ equity offering, as the two companies expand their technological collaboration to maximize the performance of Meta’s Muse personal agent app and foundation models. Meta plans to expand its…

  63. Infineon and ASMedia team up on end-to-end USB 20Gbps setup for AI cameras
    TWDigiTimes Daily

    Infineon Technologies and ASMedia have announced a partnership to speed up the adoption of 20Gbps USB, driven by demand from machine-learning cameras and other physical AI devices that need to move large amounts of data quickly. Credit…

  64. Impeaching Sara Duterte won't end her family's political ambitions
    JPNikkei Asia
  65. JIMTOF 2026 draws 94 Taiwan machine tool exhibitors
    TWDigiTimes Daily

    One of the world's three major machine tool shows, Japan's International Machine Tool Fair (JIMTOF 2026), opened this week, with 94 Taiwan companies exhibiting. The exhibitors are showcasing precision machining, key components and smart…

  66. Japan's top shippers and power producer unite on ammonia-fueled vessels
    JPNikkei Asia
  67. Anthropic IPO filing shows the US govt is trying to be frontier AI's customer and gatekeeper
    TWDigiTimes Daily

    Anthropic's IPO filing does more than outline the economics of a rapidly scaling AI company. Its risk disclosures show frontier AI developers becoming increasingly exposed to governments that act not only as regulators, but also as…

  68. The imec Report: Scaling Optical Interconnects For The AI Era
    USSemiconductor Engineering

    The explosive growth of cloud computing and AI – from generative, and reasoning, to (ultimately) agentic systems – is creating unprecedented demand for bandwidth, as the amount of data that must be moved both within and between data…

  69. Chip Industry Technical Paper Roundup: Oct. 27
    USSemiconductor Engineering

    New technical papers recently added to Semiconductor Engineering’s library: Technical Paper Research Organizations Beyond HBM-on-GPU: Thermal Design Envelope for 3D Volumetric DRAM-on-GPU Integration imec Self-Heating and Parasitic Effects…

  70. AI spending surges, Microsoft opens exit door to 60,000 employees
    TWDigiTimes Daily

    Microsoft launched a time-limited Voluntary Exit Offer (VEO) for approximately 60,000 employees in the US, Canada, the UK and India, as the software and cloud provider reshapes its workforce around rising investment in artificial…

  71. SoftBank eyes US$50B Malaysia AI data-center expansion
    TWDigiTimes Daily

    SoftBank Group founder and CEO Masayoshi Son has selected investment sites in Malaysia and plans to invest US$50 billion in technology, Malaysian Prime Minister Anwar Ibrahim said. Following investments in the US and Europe, the group is…

  72. AI data centers drive cooling, power, and modular infrastructure demand
    TWDigiTimes Daily

    As AI server computing density and server rack power consumption rise, data center construction must also take into account cooling, power distribution, energy management, and electromechanical infrastructure. In light of these trends, the…

  73. Anthropic IPO filing exposes massive compute buildout spanning TPUs, GPUs, and dedicated data centers
    TWDigiTimes Daily

    Anthropic's IPO filing reveals an infrastructure strategy extending far beyond conventional cloud procurement, as the Claude developer locks in hundreds of billions of dollars of compute capacity across Google, Amazon, Microsoft, Broadcom…

  74. Chinese brands drop soccer star Kylian Mbappe for 'offending' nation
    JPNikkei Asia
  75. Australia's Woodside to exit long-stalled Timor gas project
    JPNikkei Asia
  76. Indonesia-Singapore spats escalate over haze, MSCI
    JPNikkei Asia
  77. From enterprise data to AI agents: Building the foundation for AI at scale
    iTnews Asia

    As Southeast Asian enterprises move from AI experimentation to production, trusted data, governance and integrated applications are becoming essential to turning enterprise knowledge into measurable business value.

  78. PSMC capacity crunch holds back Jeju Semiconductor's growth
    TWDigiTimes Daily

    South Korean memory chip designer Jeju Semiconductor is struggling to secure enough wafer capacity from Taiwan's Powerchip Semiconductor Manufacturing (PSMC), limiting its sales growth as strong memory demand drives up prices and…

  79. Server CPU shipments to exceed 50 million units in 2027; AMD's server CPU revenue will surpass Intel
    TWDigiTimes Daily

    According to DIGITIMES, Agentic AI, AI server expansion, cloud service provider (CSP) in-house CPU ramp-up, and replacement demand for general-purpose servers will lift global server CPU shipments to 52.6 million units in 2027, up 45.7%…

  80. Broadcom, financial firms agree on US$63 billion funding for Anthropic chip purchases
    TWDigiTimes Daily

    Broadcom has reached a deal with several financial firms for them to finance the purchase of US$63 billion worth of its chips for Anthropic. The agreement highlights how the line between supplier and lender is becoming blurred among some…

  81. Japan struggles to attract foreign nurses amid staffing crunch
    JPNikkei Asia
  82. Mitsubishi Materials scales thermistor capacity sixfold amid AI boom
    TWDigiTimes Daily

    Mitsubishi Materials will expand production of thermistors, temperature sensors used in optical communications equipment, as the spread of generative AI drives demand for optical transceivers. Its Thailand facility will gradually add…

  83. Nikkei Asia wins gold at WAN-IFRA Asian Media Awards
    JPNikkei Asia
  84. AI boom sparks circuit board crunch across global tech supply chain
    JPNikkei Asia
  85. AI demand squeezes automotive memory, pushing automakers to take control of chip supply
    TWDigiTimes Daily

    Expanding investment in AI data centers is putting pressure on memory supply and demand across the automotive industry. As semiconductor companies prioritize high-bandwidth memory (HBM) and server products, automotive memory faces rising…

  86. China AI token debate highlights challenge of monetizing compute
    TWDigiTimes Daily

    More than half of China's AI token usage is reportedly concentrated in short dramas and video generation, but ByteDance's Doubao has disputed the figure, saying video-generation models account for only about 10% of overall computing…

  87. The Anthropic IPO filing: 14-fold 1H26 revenue growth behind a US$80.75 billion paper loss
    TWDigiTimes Daily

    Anthropic's first public filing for its IPO has opened the books on one of the world's fastest-growing AI companies, revealing extraordinary revenue growth alongside a headline net loss that substantially overstates the deterioration in…

  88. Kazakhstan positions AI infrastructure as a potential export industry after Nvidia CEO's visit
    TWDigiTimes Daily

    Nvidia CEO Jensen Huang's visit to Kazakhstan highlighted plans to build an AI computing hub for Europe, with projects spanning GPU clusters, data centers, talent, and physical AI. The initiative depends on power infrastructure, foreign…

  89. US designates helium as critical material amid Middle East supply shock
    TWDigiTimes Daily

    As military conflict in the Middle East cripples global energy infrastructure, the United States is emerging as the critical anchor for the world's high-tech manufacturing sector. Amid a historic supply crisis for helium, an indispensable…

  90. Ennostar's optical push could add new growth engine for 2027 and onward
    TWDigiTimes Daily

    Ennostar Holdings is expanding into optical communications as Micro LED and Micro VCSEL products move into customer sample-validation stages. The company expects the business to enter substantive growth in 2027, supported by potential…

  91. Fast Path, Exact Path: One Library of Hardwired Operators for Streaming and Exact Canny Edge Detection on a Low-Cost FPGA
    USarXiv cs.AR (Architecture)

    arXiv:2610.33107v1 Announce Type: new Abstract: Hysteresis edge tracking, the final stage of Canny edge detection, follows edge chains of unbounded length and therefore resists single-pass streaming; FPGA designs either approximate it or…

  92. Real-Time Recurrent Qubit-State Discrimination on a Heterogeneous AI-Engine-FPGA Accelerator
    USarXiv cs.AR (Architecture)

    arXiv:2610.33843v1 Announce Type: new Abstract: Qubit readout, or qubit state discrimination, remains one of the slowest and most error-prone operations in superconducting quantum processors, making fast, accurate, and hardware-efficient…

  93. Adapting DNNs to Embrace Voltage Scaling Induced Errors via Accumulation-Aware Quantization
    USarXiv cs.AR (Architecture)

    arXiv:2610.33824v1 Announce Type: new Abstract: The pervasive deployment of Deep Neural Networks (DNNs) across the computing spectrum has brought growing attention to energy efficiency, alongside raw performance. While voltage scaling is a…

  94. ADACT: PDE-Constrained Thermal Hardware/Software Co-Design Adjoint-Driven Optimal Control via Finite Difference Discretization
    USarXiv cs.AR (Architecture)

    arXiv:2610.35569v1 Announce Type: cross Abstract: By virtue of the sheer compute demands of large-scale artificial intelligence (AI) and high-performance computing (HPC), modern multi-core microprocessors suffer from severe power density…

  95. Double KV Cache Optimization on Apple Silicon: Block-FP4, Radial Quantization, and Fused Metal Kernels
    USarXiv cs.AR (Architecture)

    arXiv:2610.33076v1 Announce Type: new Abstract: Key-value (KV) cache capacity and memory traffic are primary bottlenecks in long-context autoregressive inference. We evaluate five cache paths on an Apple Silicon M5 Max using MLX: a native…

  96. The Analytical Investigation of Inductor-Coupled Mixers
    USarXiv cs.AR (Architecture)

    arXiv:2610.33424v1 Announce Type: new Abstract: Inductor- and transformer-coupled balanced mixers are usually analyzed with switching approximations, conversion matrices, harmonic balance, or circuit simulation. This paper instead derives…

  97. How AI Load Swings Reduce Datacenter Power Conversion Efficiency
    USarXiv cs.AR (Architecture)

    arXiv:2610.33300v1 Announce Type: new Abstract: AI training and inference workloads make GPU power swing repeatedly between compute and communication phases. As a result, the cascaded converters in a datacenter rack keep moving between…

  98. DiffRouter: A Differentiable Routing Framework for UltraScale FPGAs
    USarXiv cs.AR (Architecture)

    arXiv:2610.33510v1 Announce Type: new Abstract: Routing has become a major bottleneck in large-scale FPGA design flows, where traditional routers rely on sequential net-by-net processing and rip-up-reroute heuristics. This inherent…

  99. Toward Near-Optimal Floorplanning with Complex Design Constraints for Modern SoCs
    USarXiv cs.AR (Architecture)

    arXiv:2610.34046v1 Announce Type: new Abstract: Modern system-on-chip (SoC) floorplanning is increasingly challenged by highly coupled complex design constraints. Existing methods often relax constraints into objectives via multi-objective…

  100. SPARTA: A Sparse Ternary LLM Accelerator with Adaptive Attention and Unified Datapath for Long-Context Edge Inference
    USarXiv cs.AR (Architecture)

    arXiv:2610.34824v1 Announce Type: new Abstract: LLMs on edge devices are constrained by limited memory and bandwidth. Although 1.58-bit ternary weights greatly reduce model storage, long-context inference is increasingly bottlenecked by…

  101. Vision Transformers on the Edge: A Comprehensive Survey of Model Compression and Acceleration Strategies
    USarXiv cs.AR (Architecture)

    arXiv:2503.02891v4 Announce Type: replace-cross Abstract: In recent years, vision transformers (ViTs) have emerged as powerful and promising techniques for computer vision tasks such as image classification, object detection, and…

  102. Detection of Single Stuck-at and Bridging Faults in Cluster-based FPGA Architectures
    USarXiv cs.AR (Architecture)

    arXiv:2610.35026v1 Announce Type: new Abstract: This paper presents a new technique for detecting single stuck-at and bridging faults in cluster-based FPGA architectures, using same set of test vectors generated by a test pattern generator…

  103. Optimizing Physical Neural Networks through Device Variable Selection
    USarXiv cs.AR (Architecture)

    arXiv:2610.33522v1 Announce Type: cross Abstract: Physical neural networks (PNNs) leverage device physics in electronics, optics, and related domains to enable energy-efficient neural networks. Broken-isomorphism PNNs replace linear…

  104. EGGPlacer: Energy-Guided Generative Chip Placement with Diffusion Models
    USarXiv cs.AR (Architecture)

    arXiv:2610.33099v1 Announce Type: new Abstract: Chip placement strongly affects routed wirelength and thermal behavior, yet these downstream metrics are difficult to evaluate during macro placement. Learning-based macro placers commonly…

  105. Whole-Route Planning of a Memory Bus Between Two BGAs: An Exact Model of Crossings and Layer Changes, Negotiated to DRC-Clean Copper
    USarXiv cs.AR (Architecture)

    arXiv:2610.35601v1 Announce Type: new Abstract: Routing the bus between a processor and its DRAM remains largely manual in printed circuit board (PCB) design, and published algorithms optimize proxies such as crossing or layer counts. We…

  106. AnalogPortBench: Benchmarking LLMs on Analog Circuit Porting
    USarXiv cs.AR (Architecture)

    arXiv:2610.33088v1 Announce Type: new Abstract: Analog circuit design still relies heavily on expert knowledge, and porting proven designs to new process nodes is a recurring bottleneck. As a lower supply reduces available voltage…

  107. FADiff: Fusion-Aware Differentiable Optimization for DNN Scheduling on Tensor Accelerators
    USarXiv cs.AR (Architecture)

    arXiv:2511.22348v3 Announce Type: replace Abstract: Efficiently deploying Deep Neural Networks (DNNs) - particularly Large Language Models (LLMs) - on tensor accelerators is essential for achieving high throughput and energy efficiency at…

  108. MCF: A Memory- and Communication-Aware Co-Designed Framework for Long-Context LLM Inference
    USarXiv cs.AR (Architecture)

    arXiv:2610.34731v1 Announce Type: new Abstract: This paper presents Memory- and Communication-Aware Framework for Long-Context LLM Inference (MCF) that optimizes long-context inference through communication-aware and stage-adaptive…

  109. Realizing Low-Precision Storage and Sparse Execution in a Folded FPGA Speech Enhancer
    USarXiv cs.AR (Architecture)

    arXiv:2610.35144v1 Announce Type: cross Abstract: Hardware-aware training often optimizes FPGA resources, latency, or bit operations, but folding changes what compression saves: shared arithmetic decouples precision, weight storage, and…

  110. PaFeDSE: Pareto-Feedback-Guided Microarchitecture Design Space Exploration for Efficient Pareto Search
    USarXiv cs.AR (Architecture)

    arXiv:2610.34838v1 Announce Type: new Abstract: Efficient microarchitecture design space exploration (DSE) is critical for optimizing performance, power, and area trade-offs in modern out-of-order processors. While machine-learning-based…

  111. AMD, Nvidia target substrate capacity amid 4x expansion constraints
    TWDigiTimes Daily

    AMD and Nvidia have separately invested in supply-chain capacity expansion, making substrates, packaging, and testing key priorities as AI demand diversifies. The moves show that chipmakers are extending competition beyond performance and…

  112. Fujifilm expands South Korean capacity to meet AI chip demand
    TWDigiTimes Daily

    Fujifilm Electronic Materials is expanding its advanced semiconductor materials footprint in South Korea by increasing extreme ultraviolet (EUV) photoresist capacity in Pyeongtaek and expanding chemical mechanical polishing (CMP) slurry…

  113. FANUC raises forecast as AI, automation demand drives record revenue
    TWDigiTimes Daily

    Japanese robot maker FANUC raised its fiscal year 2026 revenue forecast on October 23 as demand for AI and automation strengthened, putting sales on track to surpass JPY1 trillion for the first time. The company also expects operating…

  114. Charts: Why German automakers are becoming marginalized in US BEV market
    TWDigiTimes Daily

    German automakers maintained overall sales in the US during the third quarter, but that momentum came mainly from gasoline-powered vehicles. In the battery electric vehicle (BEV) market, BMW, Mercedes-Benz, Volkswagen, Porsche and Audi…

  115. Nvidia says GPUs last 8 years, unlocking AI financing
    TWDigiTimes Daily

    After the GTC Berlin conference, Nvidia CEO Jensen Huang and Silicon Valley technology investor Gavin Baker spoke with Yahoo Finance for about half an hour. Huang confirmed Nvidia GPUs can remain in service for up to eight years and said…

  116. Malaysia conditionally approves first rare-earth export under new rules
    TWDigiTimes Daily

    Malaysia has approved its first export of unprocessed rare earths under a new regulatory framework, with the shipment intended for research and development, although officials did not identify the recipient. Credit: AFP

  117. Meta launches US$35 billion equity raise, funds US$60 billion at-the-market program for AI build-out
    TWDigiTimes Daily

    Meta Platforms has launched approximately US$35 billion of concurrent underwritten public offerings and plans an approximately US$60 billion at-the-market (ATM) program, as it funds capex it expects to reach US$135 billion to US$145…

  118. Applied Materials to raise Besi stake to 20% as hybrid-bonding orders surge
    TWDigiTimes Daily

    Applied Materials plans to increase its stake in Dutch semiconductor equipment maker BE Semiconductor Industries (Besi) to approximately 20% from 9%, deepening its involvement in advanced chip packaging as demand for hybrid-bonding…

  119. Nvidia's Berlin keynote highlights AI scaling challenges in Europe
    TWDigiTimes Daily

    Nvidia opened GTC 2026 in Berlin on October 21 with a keynote highlighting AI infrastructure, security architecture, and Physical AI, while a Greenpeace protest briefly interrupted proceedings. A robot demonstration also exposed limits in…

  120. Tech earnings season heats up ahead of US midterms
    JPNikkei Asia
  121. Japan's top shippers and power producer unite on ammonia vessels
    JPNikkei Asia
  122. Samsung, SK Hynix eye 2Q27 shipments of 8-layer HBM4 to Nvidia
    TWDigiTimes Daily

    Samsung Electronics and SK Hynix are targeting the second quarter of 2027 to begin volume shipments of eight-layer HBM4 memory to Nvidia, ZDNet Korea reported on October 26, as the AI chipmaker reassesses its memory strategy for…

  123. AWS, Google challenge CUDA as AMD, Intel cut AI switch costs
    TWDigiTimes Daily

    AWS has launched the open-source TorchNeuron tool, allowing developers to port PyTorch artificial intelligence (AI) applications to its custom Trainium chips without rewriting code. Google has followed with TorchTPU, while AMD and Intel…

  124. Inside SKT＆rsquo;s push to turn SK Hynix＆rsquo;s fabs into an AI business
    TWDigiTimes Daily

    South Korean telecommunications provider SK Telecom (SKT) has reportedly established Hy-TF, a task force dedicated to AI collaboration with SK Hynix, to upgrade its digital-twin and data-platform businesses and develop new B2B services…

  125. China supply-chain price hikes strain domestic demand
    TWDigiTimes Daily

    China’s semiconductor and electronics supply chains are reporting a wave of price increases, from Goke Microelectronics raising quotes for hybrid-packaged memory products to Hygon considering higher CPU and deep computing unit (DCU)…

  126. Vietnam's first fab reportedly faces doubts over 2027 target
    TWDigiTimes Daily

    Vietnam's first wafer fab may miss its 2027 trial-production target, Reuters reported on October 26, citing concerns about project execution and equipment procurement. Developer Viettel, however, says the project remains on schedule…

  127. Fine M-Tec rejects component blame as foldable market outlook remains positive
    TWDigiTimes Daily

    Fine M-Tec has rejected claims that components it supplied contributed to reported display damage in Apple's first foldable iPhone, saying the issue involves multiple components and precision mechanisms. The dispute could affect…

  128. DRAM price gains fall short as supply shrinks up to 50%
    TWDigiTimes Daily

    DRAM price growth has slowed sharply in the fourth quarter of 2026, but severe supply shortages are intensifying as manufacturers continue to restrict shipments. Industry sources expect DRAM supply from leading manufacturers to shrink by…

  129. AI data center boom risks squeezing power chip supplies for auto, electronics
    TWDigiTimes Daily

    The rapid expansion of AI data centers is driving demand beyond GPUs, memory chips, and advanced packaging. As power consumption in GPU clusters rises, data centers need not only more chips but also more efficient power delivery systems…

  130. Circle exec urges room for banks, startups in Asian stablecoin regimes
    JPNikkei Asia
  131. Weekly news roundup: iPhone Duo shipment target revised upward amid display issues
    TWDigiTimes Daily

    This week's coverage highlights shifting technology supply chains and rising demand across foldable smartphones, memory, energy storage, robotics, and AI infrastructure. Apple lifted its iPhone Duo shipment target after strong preorders…

  132. Japan sets JPY100 billion push to commercialize perovskite solar cells
    TWDigiTimes Daily

    Japan plans to provide about JPY100 billion (US$634.1 million) over four years to support mass production of next-generation perovskite solar cells, seeking to establish a domestic manufacturing base before China further widens its lead in…

  133. HCLTech and Rapidus open India centers to support semiconductor engineering
    TWDigiTimes Daily

    HCLTech and Japan's Rapidus have inaugurated development centers in Bengaluru and Hubli, India, expanding their semiconductor engineering collaboration across Asia. The facilities will support advanced chip work, while a planned talent…

  134. India's smartphone shipments hit a decade low in 3Q26 amid rising prices
    TWDigiTimes Daily

    India's smartphone shipments fell 20% year over year in the third quarter of 2026, reaching a decade low as higher memory costs pushed device prices up. The downturn highlights affordability pressures for consumers and could prompt…

  135. As iPhone Duo inner screens crack, Samsung capitalizes on foldable surge
    TWDigiTimes Daily

    Days after the launch of Apple's US$2,000 iPhone Duo, a wave of reports regarding broken inner screens has surfaced across Reddit, TechCrunch , and Forbes . Early buyers reported hearing a loud crunching noise before severe cracking…

  136. Intel to acquire quantum computing software firm Classiq
    TWDigiTimes Daily

    Intel has entered into a definitive agreement to acquire Classiq, a quantum computing software company, in a move aimed at combining Classiq's quantum algorithm synthesis platform with Intel's silicon-based quantum hardware and…

  137. Anthropic's IPO filing puts TSMC, HBM, direct leasing at the center of AI compute risk
    TWDigiTimes Daily

    In its preliminary IPO registration statement filed with the US Securities and Exchange Commission, Anthropic describes how much of its capacity depends on a small group of chip and memory suppliers, most of them in East Asia. Credit…

  138. Tamil Nadu adds independent chip subsidy route as Indian states vie for investment
    TWDigiTimes Daily

    India's semiconductor drive has largely been led by New Delhi, which selects projects for large central incentives while state governments add their own packages of land, infrastructure and financial support. Tamil Nadu is now giving…

  139. OPTOELEC seeks AI data center orders as Himax advances CPO optics
    TWDigiTimes Daily

    South Korean optics maker Optoelec is seeing growing interest from AI data center customers as it seeks to expand its wafer-level optics (WLO) business into high-speed optical communications, a market where Taiwanese suppliers are already…

  140. NT1 plans to assemble Chinese delivery vans in Korea in 2027
    TWDigiTimes Daily

    South Korean autonomous delivery company NT1 plans to begin assembling Chinese-designed delivery vans locally in 2027, gradually adding Korean-made components as it seeks approval to operate the vehicles on public roads. Leon, NT1…

  141. Hyundai supplier SL eyes first overseas robot deployment in US
    TWDigiTimes Daily

    South Korean automotive supplier SL is considering deploying its mobile dual-arm robot at a US factory in 2027 or 2028, potentially marking the system's first overseas deployment, a company representative said in an interview with…

  142. Tata Communications plans up to US$450 million fibre network linking India's AI hubs
    TWDigiTimes Daily

    Tata Communications announced on October 26 that it plans to invest US$400-450 million in approximately 10,000 kilometers of long-haul fiber across India, linking data centers and AI hubs as it expands its connectivity business. The…

  143. KAR bets on fleet monitoring as South Korea expands robotaxi trials
    TWDigiTimes Daily

    South Korean vehicle engineering company KAR plans to develop a monitoring system for the electronic controls it installs in autonomous vehicles, seeking recurring maintenance revenue as the country expands robotaxi trials. Amy Choi…

  144. ByteDance denies AI micro-dramas consume over half of China's AI computing power
    TWDigiTimes Daily

    The claim that more than half of China's artificial intelligence (AI) computing power is being used to produce micro-dramas has reignited debate in China's technology sector over whether the country is directing too much of its AI…

  145. Amkor turns to customer prepayments as global packaging investment accelerates
    TWDigiTimes Daily

    Amkor Technology is using customer prepayments and longer-term commitments to support an expanding manufacturing network across the US, South Korea, and Vietnam, as demand for advanced packaging outpaces the consumer electronics recovery…

  146. Microsoft offers buyouts to 60,000 employees as AI spending mounts
    JPNikkei Asia
  147. Cadence bets on its own chip design AI models as Synopsys teams up with OpenAI
    TWDigiTimes Daily

    Cadence Design Systems is betting that chipmakers will want AI models trained on its own design tools and run wherever customers choose. That puts it on a different route from rival Synopsys, which is building its specialized model with…

  148. Cadence says chip design demand is broadening beyond AI as analysts test margins, rival targets
    TWDigiTimes Daily

    Cadence Design Systems used its third-quarter earnings call to argue that demand for chip design tools is no longer tied only to the AI hardware buildout. It pointed to analog and memory chipmakers, system companies, and new entrants such…

  149. India sets October 2028 target for homegrown smartphone brand
    TWDigiTimes Daily

    India's electronics and IT minister, Ashwini Vaishnaw, has set October 23, 2028, as the date by which the country will have mobile phones that are designed, manufactured, and branded domestically, as New Delhi seeks to move beyond…

  150. Celestica's AI growth puts capacity in focus as revenue outlook reaches US$40 billion
    TWDigiTimes Daily

    Celestica's latest results suggest that its next challenge will be expanding production fast enough to serve AI infrastructure customers while keeping cash requirements under control. The company forecast US$40 billion in revenue for 2027…