News from November 2026
678 stories of semiconductor news published on the Hub in November 2026, newest first and grouped by day (UTC). Each headline opens the story's page on the Hub, with a summary and a link to the original report.
Showing 201 to 250 of 678 stories.
Monday 9 November 202697 stories
The Optical Super Highway For AIMarkets & Policy · Semiconductor Engineering · Summary
Exponentially increasing AI compute performance requires exponential interconnect speeds. The post The Optical Super Highway For AI appeared first on Semiconductor Engineering.
Memory In Physical AIMarkets & Policy · Semiconductor Engineering · Summary
Image: Semiconductor Engineering (opens in a new tab) LPDDR6 becomes an essential ingredient as intelligence moves into autonomous and mobile edge systems. The post Memory In Physical AI appeared first on Semiconductor Engineering.
NVIDIA Announces New Planned Investments in Korean Startups at the MSIT-NVIDIA AI Startup Partnership ForumMarkets & Policy · NVIDIA · Summary
Image: NVIDIA (opens in a new tab) Korea’s Ministry of Science and ICT and NVIDIA hosted the MSIT-NVIDIA AI Startup Partnership Forum at COEX. Ryu Je-myung, vice minister of science and ICT, joined NVIDIA leaders to meet with over 20 Korean startups across foundation models, physical AI, AI infrastructure and AI semiconductors.
NVIDIA also announced an investment in Korean startup Point2, with […]
Samtec AcceleRate mP connectors cut PCB spaceMarkets & Policy · eeNews Europe · Summary
Image: eeNews Europe (opens in a new tab) Samtec has expanded its AcceleRate mP connector family with new 10 mm stack-height versions that combine high-speed digital signaling and power delivery in a single board-to-board connector. The UDM6 and UDF6 connectors are designed to reduce PCB space requirements while supporting high current and data rates of up to 64 Gbps PAM4.
The development is […] The post Samtec AcceleRate mP connectors cut PCB space appeared first on eeNews Europe.
Nvidia probe of CEO unlikely to disrupt Taiwan AI supply chain: TIER headMarkets & Policy · Focus Taiwan (CNA) · Summary
Image: Focus Taiwan (CNA) (opens in a new tab) Taipei, Nov. 9 (CNA) An investigation launched by Nvidia into its founder and CEO Jensen Huang (黃仁勳) over his affair with an alleged Chinese spy is unlikely to have a major impact on Taiwan's artificial intelligence supply chain, according to the head of a Taiwanese think tank.
GF to run 7nm process next yearFoundry & Manufacturing · Electronics Weekly · Summary
Image: Electronics Weekly (opens in a new tab) Globalfoundries says it intends to demo a 7nm FDSOI process in early 2027 called FDX Fusion. The process claims to deliver more than double the transistor density of first-generation 22FDX, […] The post GF to run 7nm process next year appeared first on Electronics Weekly.
TSMC rumoured to be eyeing 3D DRAMFoundry & Manufacturing · Electronics Weekly · Summary
Image: Electronics Weekly (opens in a new tab) TSMC is said to have appointed vp Hsu Kuo-Chin to look at how the company could get into 3D DRAM according to multiple industry reports, but it would seem that […] The post TSMC rumoured to be eyeing 3D DRAM appeared first on Electronics Weekly.
Jensen Huang apologises: Nvidia board investigatesMarkets & Policy · Electronics Weekly · Summary
Image: Electronics Weekly (opens in a new tab) After Jensen Huang (pictured) admitted an affair with Alice Huang, a senior exec at Singapore datacentre company Megaspeed, who U.S. intelligence officials believe is a Chinese spy, the board of […] The post Jensen Huang apologises: Nvidia board investigates appeared first on Electronics Weekly.
DeepX Tapes Out Second-Gen Chip, Projects 6.7× Performance Gain With LPDDR5x-PIMDesign & IP · EE Times · Summary
Image: EE Times (opens in a new tab) DeepX is targeting edge markets including robotics with its second-generation AI accelerator, which can run large LLMs on device in under 5 W. The post DeepX Tapes Out Second-Gen Chip, Projects 6.7× Performance Gain With LPDDR5x-PIM appeared first on EE Times.
NVIDIA, Rebellions and SK Telecom Announce Collaboration to Advance Sovereign AI Infrastructure With NVIDIA NVLink FusionDesign & IP · NVIDIA · Summary
Image: NVIDIA (opens in a new tab) NVIDIA, Rebellions and SK Telecom today announced a collaboration to advance Korea’s sovereign AI infrastructure — using NVIDIA NVLink Fusion™ to bring together NVIDIA’s open infrastructure platform, Rebellions’ Korean-designed AI processors and SK Telecom’s AI data center, model and service capabilities.
NVIDIA Dynamo Recipes Bring Korea’s Leading Open Models to ProductionDesign & IP · NVIDIA · Summary
Image: NVIDIA (opens in a new tab) NVIDIA is releasing Dynamo-optimized deployment recipes for Korea’s leading open models: A.X K2 from SK Telecom, K-EXAONE 2.0 from LG AI Research, Solar Open 2 from Upstage and Motif-3 from Motif Technologies.
As models grow beyond the capacity of a single GPU, they must be split across multiple GPUs, requiring careful coordination of request routing […]
From Chips to Skills, NVIDIA and Korea Expand Sovereign AI CollaborationMarkets & Policy · NVIDIA · Summary
Image: NVIDIA (opens in a new tab) NVIDIA and Korea’s Ministry of Science and ICT (MSIT) are expanding their collaboration across the country’s sovereign AI ecosystem, from AI semiconductors and data center infrastructure to talent, startups, models, agents and security.
The collaboration lays the groundwork for Korean NPU and XPU companies to explore participation in the NVIDIA NVLink Fusion ecosystem, and includes […]
Korea’s AI Ecosystem Advances and Secures Agentic AI With NVIDIAMarkets & Policy · NVIDIA · Summary
Image: NVIDIA (opens in a new tab) As AI agents take on more work, they need capable open models and security that advances alongside them.
Hyundai Motor Group is exploring NVIDIA NemoClaw blueprints and the NVIDIA OpenShell runtime, while AIM Intelligence is integrating its Starfort platform with NVIDIA OpenShell to support secure AI agents for employee productivity and enterprise workflows. A key […]
Korea Advances Physical AI Across Industries With NVIDIAEDA & Tools · NVIDIA · Summary
Image: NVIDIA (opens in a new tab) Tapping into the country’s strengths in manufacturing and engineering, NVIDIA is helping advance Korea’s physical AI ecosystem. The country’s industry leaders are applying accelerated computing, AI and simulation to everything from semiconductor manufacturing to robotics and autonomous vehicles.
These efforts are helping bring physical AI from development into real-world production. Samsung Advances Semiconductor Engineering With […]
BitDecoding-FP4: Native FP4 KV Cache Decoding via Algorithm-System Co-DesignMarkets & Policy · arXiv cs.AR (Hardware Architecture) · Summary
The key-value (KV) cache dominates memory capacity and bandwidth in long-context LLM serving.
While 4-bit microscaling formats (NVFP4, MXFP4) shrink the cache and run attention on FP4 Tensor Cores, a fundamental gap remains between algorithmic quantization and system-aware acceleration: existing FP4 KV systems use vanilla quantization that causes severe accuracy drops, whereas accuracy-oriented 4-bit algorithms rely on vector codebooks whose lookup overhead degrades decoding throughput.
We pres…
Atik: A Shared-Mesh Transformer Accelerator for RISC-VDesign & IP · arXiv cs.AR (Hardware Architecture) · Summary
Transformer inference repeatedly invokes dense matrix multiplication and attention, including its data-dependent softmax normalization. We present Atik, an open-source RoCC accelerator for RocketChip that executes matrix multiplication and dense attention through one shared fixed-point MAC mesh.
Atik accepts and returns BF16 tensors, uses a streaming online-softmax schedule that never materializes the full score matrix, and is invoked through a small descriptor-based ABI. We evaluate three mesh…
Robust & Scalable Website Fingerprinting via Minimal DRAM TrafficMarkets & Policy · arXiv cs.AR (Hardware Architecture) · Summary
Website fingerprinting attacks infer websites visited by users and threaten their privacy. These attacks are classified as on-path, which monitor network traffic, and co-located, which exploit shared system resources.
Existing co-located attacks, such as sweep-counting and loop-counting, achieve over 90% accuracy but suffer significant performance degradation on many-core systems. Loop-counting accuracy drops from 93% on 8 cores to 50% on 24 cores, while sweep-counting accuracy decreases from 8…
Gain-Dependent Entropy Characterization of a Zener Avalanche Noise TRNG via Dual NIST Standards: Toward Low-Cost Hardware Entropy for ICS DeploymentsMarkets & Policy · arXiv cs.AR (Hardware Architecture) · Summary
True random number generators based on Zener avalanche noise offer a low-cost entropy source for industrial control system and IoT edge devices, yet the relationship between amplifier gain, coupling capacitance, and entropy quality remains uncharacterized under current NIST standards.
This paper presents a comprehensive entropy assessment of an op-amp-based Zener avalanche noise generator across five gain configurations and two coupling capacitor values, evaluated against the NIST SP 800-22 sta…
CPU Micro-Architectural Characteristic Sensitivity Study on CPU-Side Agentic Task LatencyDesign & IP · arXiv cs.AR (Hardware Architecture) · Summary
We perform a novel study of CPU micro-architectural characteristics on how they affect agentic AI CPU side latency.
Small enterprises and individual users looking to reduce costs and seeking data privacy benefit from running agents locally, motivating an investigation into the optimal performance characteristics of an edge CPU to improve user experience.
We investigate the relationship between CPU core frequency, CPU uncore frequency, active physical cores, and symmetric multi threading with ag…
Memory as a Temporal Resource: Lifetime Contracts for Multi-Retention SystemsMarkets & Policy · arXiv cs.AR (Hardware Architecture) · Summary
Computers move and rewrite data whose useful lifetimes range from inner-loop iterations to years, yet memory systems usually hide retention as a fixed physical property. We argue that required lifetime should become a first-class resource across algorithms, programming systems, controllers, and devices.
Building on prior lifetime-aware placement, we propose an event-based validity contract that links application completion events to retention-class selection and renewal of protection. Applicati…
TwinHash: Two-Sided Object Identity Management for Time- and Energy-Efficient Agentic WorkflowsMarkets & Policy · arXiv cs.AR (Hardware Architecture) · Summary
Agentic artificial intelligence (AI) services interleave large language model (LLM) generation with external tool execution, and increasingly rely on search-based workflows that repeatedly revisit shared reasoning states and computations.
These patterns create substantial redundancy across tool outputs, LLM responses, and key-value (KV) prefixes, yet existing serving systems primarily optimize shared prompt prefixes, while naive input-keyed caching is unsafe when identical requests produce diff…
Enabling Retraining-Free Accurate Inference in Analog Compute in Memory AcceleratorsDesign & IP · arXiv cs.AR (Hardware Architecture) · Summary
Analog compute-in-memory (CIM) accelerators compute matrix-vector products directly inside memory arrays, eliminating weight movement and promising order-of-magnitude gains in inference efficiency. However, converter and device errors make pretrained models unusable on these arrays, and hardware-aware retraining is costly.
We enable retraining-free, accurate inference of pretrained models on simple analog arrays with low-cost, shared calibration. Our key observation is that analog error depends…
SAFTAC: Simulation-Augmented Fine-Tuning of Open-Source LLMs for Analog Circuit DesignEDA & Tools · arXiv cs.AR (Hardware Architecture) · Summary
Circuit design for analog integrated circuits (ICs) is challenging because circuit topology, device sizing, and circuit performance are tightly coupled.
Existing large language model (LLM)-based methods either rely on inference-time correction around frozen models or train on limited circuit types without directly using simulation outcomes as training signals.
A key bottleneck is that existing datasets do not provide sufficient task information for training LLMs on end-to-end specification-cond…
RoadBlock: Rethinking GPU Tensor Core Microarchitecture for Emerging Microscaling Format SupportDesign & IP · arXiv cs.AR (Hardware Architecture) · Summary
While multiple recent custom Microscaling (MX) format proposals preserve higher model quality over OCP MX, their hardware evaluations remain confined to standalone MAC-unit prototypes, ignoring the system-level integration overheads inside real GPU Tensor Cores.
To this end, we present RoadBlock, an open-source framework for accurate end-to-end modeling of custom MX formats in a programmable Tensor Core, built on the RISC-V Vortex GPGPU. RoadBlock exposes data size, scale-factor size, and block…
Fibocom Showcases U6G-Capable FWA and PCBA-Level Solutions at MWC Doha 2026Markets & Policy · PR Newswire · Summary
Image: PR Newswire (opens in a new tab) DOHA, Qatar, Nov. 8, 2026 /PRNewswire/ -- At MWC Doha 2026, Fibocom is showcasing its 5G and 5G-Advanced broadband portfolio, featuring FG390, FG370 and FG670 modules alongside PCBA-level solutions for indoor units (IDUs), outdoor units (ODUs), MiFi devices and 5G dongles. The portfolio...
Nvidia CEO Jensen Huang starring in possible remake of The Spy Who Shagged MeDesign & IP · The Register · Summary
Image: The Register (opens in a new tab) Admits to affair after company board opens probe into alleged contact with Chinese asset who bought GPUs in bulk
Huawei says it’s made a 100 percent Chinese PCMarkets & Policy · The Register · Summary
Image: The Register (opens in a new tab) PLUS: Japan declares infosec emergency; Apple’s makes biggest school Mac sale, ever, anywhere, in the Philippines; And more!
Nvidia CEO Jensen Huang admits to affair; board orders investigationMarkets & Policy · South China Morning Post - Tech · Summary
Nvidia’s board has announced an independent investigation as chief executive Jensen Huang said in a memo to employees that he had been unfaithful to his wife. Huang admitted to an extramarital affair and apologised to his family and colleagues in the memo, which the company published on its website on Sunday.
“I was unfaithful to my wife. My actions hurt her and our children.
There is no excuse for what I did. I am deeply sorry,” Huang wrote.
He apologised and pledged to cooperate fully with th…
Dell PowerEdge R770AP Review One Fast Intel Xeon 6 ServerDesign & IP · ServeTheHome · Summary
Image: ServeTheHome (opens in a new tab) Today we are taking a look at Dell's PowerEdge R770AP. The 2U server goes all-in on performance with a dual-socket Xeon 6 setup and 24 channels of DDR5 memory The post Dell PowerEdge R770AP Review One Fast Intel Xeon 6 Server appeared first on ServeTheHome.
Sunday 8 November 202616 stories
Nvidia investigates CEO Jensen Huang's relationship with alleged Chinese spyMarkets & Policy · CNBC Technology · Summary
Image: CNBC Technology (opens in a new tab) Huang told Nvidia staff in an email that he had an affair and that the company's board has opened an investigation into allegations the woman is a Chinese spy.
Nvidia CEO Jensen Huang admits to having an affair with suspected Chinese spyMarkets & Policy · Tom's Hardware · Summary
Image: Tom's Hardware (opens in a new tab) Nvidia's board of directors is investigating the report about Jensen Huang's alleged affair with a client, claims its confidence in Jensen and leadership team.
CEO Interview with Mark Rushworth of FinchettoMarkets & Policy · SemiWiki Forums · Summary
Mark Rushworth is CEO of the all-optical network switch company, Finchetto, a photonics startup developing fully optical network switching technology for AI and hyperscale data centers.
He specializes in integrated photonics and optical networking, with a focus on reducing the power consumption and latency of modern computing… Read More The post CEO Interview with Mark Rushworth of Finchetto appeared first on SemiWiki.
Statement from NVIDIA Board of DirectorsMarkets & Policy · NVIDIA · Summary
The NVIDIA Board is initiating an independent investigation in light of recent reporting that founder and CEO Jensen Huang has been the target of a foreign intelligence effort. A committee of ...
Nvidia CEO Jensen Huang faces bipartisan congressional scrutiny over reported China spy linkMarkets & Policy · CNBC Technology · Summary
Image: CNBC Technology (opens in a new tab) Lawmakers want to question Nvidia CEO Jensen Huang following reports about his relationship with an alleged Chinese intelligence asset.
CEO Interview with Albert Liu of KneronMarkets & Policy · SemiWiki Forums · Summary
Albert Liu founded Kneron in 2015 with a mission to bring on-device Edge AI and machine learning to mass-market devices and usher in the age of AI everywhere. His vision for Kneron is to realize the Edge AI Net, democratizing AI by moving inferencing away from the centralized cloud “brain” and onto devices everywhere.
Liu previously… Read More The post CEO Interview with Albert Liu of Kneron appeared first on SemiWiki.
Diamond Foundry creates world's first 200mm single-crystal diamond waferFoundry & Manufacturing · Tom's Hardware · Summary
Image: Tom's Hardware (opens in a new tab) This 200mm single-crystal diamond wafer is designed to efficiently remove heat and prevent hotspots from forming on AI processors.
Nvidia released its 8800 GTX GPU 20 years ago today, celebrates with Steam game homageDesign & IP · Tom's Hardware · Summary
Image: Tom's Hardware (opens in a new tab) On this day in 2006, Nvidia released the legendary GeForce 8800 GTX graphics card and it is celebrating the event with a very demanding (but free) interactive Steam demo.
The end of an era as Linux 7.5 purges the last 30-year-old Intel MMX optimizations from Pentium ageMarkets & Policy · Tom's Hardware · Summary
Image: Tom's Hardware (opens in a new tab) Linux developers are preparing to make Intel’s ancient MultiMedia eXtensions irrelevant, with the last MMX code now set to be removed from the kernel.
CEO Interview with Jun Yuan of Arcas-TechEDA & Tools · SemiWiki Forums · Summary
JUN YUAN BIO ◆Engineer for chip design, simulation, and P&R at AMD; Senior Engineer for chip verification (emulation, equivalence checking, formal verification) at Motorola since 1992. ◆Held R&D and technical management roles in formal verification EDA at Verplex, Jasper, and Cadence since 2003. ◆Founded Arcas-tech… Read More The post CEO Interview with Jun Yuan of Arcas-Tech appeared first on SemiWiki.
Two Chinese companies cleared to buy banned Nvidia AI chips have empty offices and unknown addressesDesign & IP · Tom's Hardware · Summary
Image: Tom's Hardware (opens in a new tab) The offices for New Matrix and Maginfra, both approved to receive H200 chips, appear abandoned, and people in the buildings say they're unfamiliar with the companies or have never seen anyone working there.
Open-source GPU task manager reveals exactly what's eating your VRAMDesign & IP · Tom's Hardware · Summary
Image: Tom's Hardware (opens in a new tab) A new open-source GitHub project called gputaskman has been released by Microsoft VP Scott Hanselman.
CEO Interview with Pratap Ranade of Arena PhysicaMarkets & Policy · SemiWiki Forums · Summary
Pratap Ranade is CEO and Co-Founder of Arena Physica. Pratap first trained as a physicist at Stanford and Columbia before joining McKinsey as an Associate Partner advising tech and semiconductor companies.
He then founded Kimono Labs, backed by Sam Altman, Peter Thiel and Y-Combinator, which was acquired by Palantir. There,… Read More The post CEO Interview with Pratap Ranade of Arena Physica appeared first on SemiWiki.
Homelabber uses second-gen Optane memory sticks to run a 2.8-trillion-parameter AI model on CPU aloneDesign & IP · Tom's Hardware · Summary
Image: Tom's Hardware (opens in a new tab) A Reddit homelab user ran Moonshot AI’s 2.8-trillion-parameter Kimi K3 CPU-only on Optane PMem 200 at 0.85 tokens/s with a 1-bit quant.
PC's with 64GB of RAM plummet 18% as memory prices soarMarkets & Policy · Tom's Hardware · Summary
Image: Tom's Hardware (opens in a new tab) CPU-Z data reveals PC memory capacity is dropping for the first time ever. Driven by skyrocketing AI DRAM prices, 64GB RAM adoption plummeted 18% YoY.
WoW: Forever launch used 70,480 CPU cores and 251 terabytes of RAMDesign & IP · Tom's Hardware · Summary
Image: Tom's Hardware (opens in a new tab) Inside the astonishing hardware behind World of Warcraft: Forever. 251TB of RAM and 70K cores kept millions of players online without a hitch.
Saturday 7 November 202611 stories
The Midterms Hit AI: Data Center Pushback, a Market Whipsaw, and Earnings That Tell a Different StoryMarkets & Policy · The Six Five · Summary
A midterm election that put data centers on the ballot, followed by one of the sharpest two-day market swings in recent memory, changed the AI conversation this week, even as earnings from AMD, Arm, Lattice, and Coherent pointed to demand that hasn't slowed.
Patrick Moorhead and Daniel Newman unpack what the election results could mean for the AI buildout, why the market reacted the way it did, and what the week's earnings actually showed. The handpicked topics for this week are: Data Centers L…
This RTX 5060 gaming PC costs less than the Steam Machine — grab the iBuyPower Slate prebuilt PC with massive $500 discountDesign & IP · Tom's Hardware · Summary
Image: Tom's Hardware (opens in a new tab) Best Buy knocks $500 off the iBUYPOWER Slate, bringing this RTX 5060-powered gaming desktop down to $899 with a Ryzen 5 processor, 16GB of DDR5 memory, and a 1TB SSD
Meta chief strategy officer Dave Wehner leaves to join AI infrastructure startup Fireworks as CFOMarkets & Policy · CNBC Technology · Summary
Image: CNBC Technology (opens in a new tab) Dave Wehner, Meta’s chief strategy officer, is leaving the company to become CFO of Fireworks, the AI infrastructure startup valued at $17.5 billion in July.
ASRock Rack DMRD16-2L+ Spotted with 16 DDR5 Slots for Diamond RapidsMarkets & Policy · ServeTheHome · Summary
Image: ServeTheHome (opens in a new tab) We check out the ASRock Rack DMRD16-2L+ which is a 16-channel Intel Diamond Rapids Xeon motherboard with 128x PCIe Gen6 lanes over MCIO The post ASRock Rack DMRD16-2L+ Spotted with 16 DDR5 Slots for Diamond Rapids appeared first on ServeTheHome.
TeamGroup GA Pro 2TB SSD Review: A faltering attempt at a mid-range PCIe 5.0 SSDMarkets & Policy · Tom's Hardware · Summary
Image: Tom's Hardware (opens in a new tab) The TeamGroup GA Pro mostly misses. Decent QD8 and steady-state writes can’t make up for weak overall performance, poor efficiency, and high temps.