News from November 2026

Recent generations of high bandwidth memory (HBM), such as HBM4, allow the base die of the memory stack to be manufactured using a logic process that can accommodate more complex controller functions next to the DRAM.

We present the first systematic quantitative study of function-level controller placement between the host and the HBM4 base die on a silicon interposer. We divide the controller into its main functions and evaluate four placements, from a complete host controller to a complete ba…

Machine-learning-based hardware prefetchers can learn complex access patterns, but practical deployment demands online adaptation, low-latency inference, and modest hardware cost. We present Augur, a hardware delta prefetcher based on next-generation reservoir computing (NGRC) that meets all three requirements.

Augur formulates delta prediction as a history-based regression problem. It constructs a fixed-size reservoir from hashed delta features over a short context window and predicts the next…

Modern software uses indirect branches for various purposes including, but not limited to, virtual method dispatch and implementation of switch statements.

Because an indirect branch's target address cannot be determined prior to execution, high-performance processors depend on highly-accurate indirect branch prediction techniques to mitigate control hazards. This paper proposes a new indirect branch prediction scheme that predicts target addresses at the bit level.

Using a series of perceptron…

Ternary large language models (tLLMs) replace general multiplication in dominant linear layers with sign-controlled accumulation, shifting hardware bottlenecks toward packed-weight delivery, reduction, and operator occupancy. PENSA is an end-to-end FPGA accelerator for the BitNet b1.58 2B4T model on an AMD Alveo U50.

Two row-partitioned ternary matrix engines consume packed TQ1_0 weights from eight independent high-bandwidth memory (HBM) pseudo-channels, each processing 1,280 ternary weights pe…

Embedded AI is moving from perception to prediction. Robots now use world models$\unicode{x2014}$learned simulators that predict the outcome of candidate actions in a compressed latent space before any is executed.

Every predicted step must be computed on board within a tight power and thermal budget, yet the energy these steps fundamentally require, and how far real hardware sits from it, have not been characterized. We derive lower bounds on the arithmetic operations and the CMOS dynamic ener…

How connected inspection, metrology and software can help manufacturers move SiPh and CPO toward high-volume production. The post Overcoming Silicon Photonics Process Control Challenges in the AI Era appeared first on Semiconductor Engineering.

Existing inspection tools struggle to capture defects, especially those hidden at the bond interface. The post Shrinking Hybrid Bonding Pitches Push Inspection Limits appeared first on Semiconductor Engineering.

Changing test strategies put a premium on how closely the design ecosystem and the test ecosystem line up. The post Streaming Scan Network Test Over PCIe appeared first on Semiconductor Engineering.

A comprehensive guide to expanding materials research capabilities. The post The Definitive AFM Modes Handbook: 2026 Edition appeared first on Semiconductor Engineering.

AI is reshaping the tools, data, and collaboration models used to design, manufacture, and test silicon. The post AI’s Efficiency Challenge For The Semiconductor Industry appeared first on Semiconductor Engineering.

Manufacturers are using test data and predictive models to identify marginal chips and determine when additional screening is justified. The post Which Passing Chips Need More Testing? appeared first on Semiconductor Engineering.

Rapidly changing workloads, more test insertions, and fragmented data complicate chip testing. The post Test Struggles To Keep Pace With AI appeared first on Semiconductor Engineering.

How yield prediction technology can reduce trial-and-error in advanced packaging lithography. The post Reducing Lithography Learning Cycles To Accelerate FOPLP Ramps appeared first on Semiconductor Engineering.

Rapidus, Japan’s state-backed effort to make 2nm chips, has failed to make a functioning 2nm SRAM reports Bloomberg. The failure puts in question its targets of entering 2nm mass production […] The post Delays reported at Rapidus appeared first on Electronics Weekly.

TAIPEI, Nov. 10, 2026 /PRNewswire/ -- ASE Technology Holding Co., Ltd. (NYSE: ASX, TAIEX: 3711, "ASEH" or the "Company"), announces its unaudited consolidated net revenues for October 2026. CONSOLIDATED NET REVENUES (UNAUDITED) Oct Sept Oct Sequential YoY (NT$ Million) 2026 2026 2025...

Chang Ming delivers scarce semiconductors, with rigorous 9-step quality inspection & global supply chain support for European & American electronics buyers. The post We Source Chips Others Cannot Secure, 2026 appeared first on EE Times.

[The content of this article has been produced by our advertising partner.] Huawei brought together nearly 100 industry experts and ecosystem partners at the HDD (Huawei Developer Day) HarmonyOS Ecosystem Executive Summit in Hong Kong on November 9, 2026, as regional enterprises increasingly collaborate to meet the surging demand for seamless outbound travel and lifestyle experiences across the Guangdong-Hong Kong-Macao Greater Bay Area (GBA).

Held under the theme "Open Innovation · Mutual...

At electronica 2026, Inova Semiconductors is showcasing a development platform for humanoid robots built on Inova’s APXpress high-speed networking technology and low-power Lattice CertusPro and Lattice CrossLink NX FPGAs. DLR’s […] The post Electronica: Development platform for humanoid robots appeared first on Electronics Weekly.

Infineon is processing 300mm SiC wafers into functional power devices at its fab in Villach, Austria. Infineon used existing silicon (Si) manufacturing equipment.

The larger wafer diameter offers 2.3 times […] The post Infineon first to fab SiC on 300mm wafers appeared first on Electronics Weekly.

A private equity company dedicated to making investments in the semiconductor supply chain has raised €1 billion The company, Ardian Semiconductor, is part of the French private equity company Ardian. […] The post Ardian Semi raises €1bn appeared first on Electronics Weekly.

Gallium oxide (Ga₂O₃) is moving into early production mode. The ultra-wide bandgap material with a bandgap around 4.7 to 4.9 eV — nearly fifty percent wider than SiC — is […] The post Gallium Oxide wafers in early production appeared first on Electronics Weekly.

GigaDevice officially lunches its new-generation GD32H7 MCU portfolio built on a unified chip architecture. Discover More.

The post GigaDevice New-Gen High-Performance MCU Portfolio for Full-Scenario Industrial Applications appeared first on EE Times.

Delft-based Delmic is introducing Aurora, a system that automatically detects crystal defects in semiconductor materials.

TSMC reports its monthly sales figures for October on Tuesday.

Two mainland Chinese chip suppliers made sharply divergent trading debuts in Hong Kong on Tuesday after raising a combined HK$2.2 billion (US$280 million), underscoring investors’ selective appetite for semiconductor stocks despite a pickup in new listings.

Suzhou Denglin Technology, the week’s largest listing, fell by as much as 8.6 per cent in morning trade after raising HK$1.4 billion in its initial public offering (IPO). The graphics processing unit maker priced its shares at HK$231 – the...

Prices for Nvidia graphics processing units (GPUs) are falling in China’s largest electronics market, as traders shrug off the frenzy of market speculation over tightening export controls in the wake of the US midterm elections and the Jensen Huang affair.

The US chip giant’s top-of-the-range GeForce RTX 5090 graphics card surged in price last month at Shenzhen’s Huaqiangbei market, but that was mostly driven by rumours that Nvidia was going to halt production to focus on high-end enterprise...

Infineon Technologies AG has successfully processed 300mm silicon carbide wafers into functional power devices at its fab in Villach, Austria. The post Infineon Processes Functional Power Devices on 300mm SiC Wafers appeared first on Semiconductor Digest.

New global competition invites university students to develop research and solutions addressing real-world semiconductor challenges. The post SEMI Foundation Launches Global Chip the Future Challenge appeared first on Semiconductor Digest.

At SEMICON Europa 2026, the Fraunhofer Institute for Photonic Microsystems IPMS will showcase current research results and technological developments for the semiconductor industry. The post Fraunhofer IPMS Highlights Semiconductor Technology Developments at SEMICON Europa appeared first on Semiconductor Digest.

Over the past decade, Kubernetes established itself as the standard for orchestrating microservices and containerized applications. Today, we're witnessing the next major architectural shift in distributed computing: AI workloads moving from isolated research prototypes into large-scale, multi-tenant production fleets.

In this post, we'll walk through the open-source stack that powers this shift and show you the sample manifests that you can try on your own Kubernetes clusters. You can also joi…

Class-action lawsuit and legal investigations against Nvidia over securities fraud start flying

Yesterday (U.S. Central time) I appeared on BBC News Business Today to discuss the unfolding saga of NVIDIA CEO Jensen Huang and its possible implications for the company as well as trade relations between the U.S. and China.

No video is available online, but you can read on for an AI-generated, human-edited overview and summary […] The post Patrick Moorhead Discusses Jensen Huang and NVIDIA on the BBC, November 8, 2026 — Broadcast Analysis appeared first on Moor Insights & Strategy.

Order marks an important step on the industry’s path from pilot lines to volume manufacturing of glass core substrates for AI and high-performance computing. The post SCHMID to Supply Equipment for Advanced Packaging Facility in South Asia appeared first on Semiconductor Digest.

EV Group today announced that it is supporting quantum chip developers and manufacturers in addressing the complex manufacturing requirements associated with scaling quantum computing systems. The post EV Group Advances Qubit Scaling Through Semiconductor Packaging Technology appeared first on Semiconductor Digest.

Mixx Technologies, Inc. today set out its view of high-density optical scale-up as a member of the Expanded Beam Optics Multi-Source Agreement (EBO MSA) and a contributing member of the Open CPX MSA. The post Mixx Technologies Backs Open Standards for Near-Package Optical Connectivity appeared first on Semiconductor Digest.

HTEC research finds semiconductor companies are ahead in scaling AI, yet fragmented measurement is making it harder to connect engineering gains to business outcomes. The post Semiconductor Companies Get Tough on AI Investment, but Measuring Returns Remains a Challenge appeared first on Semiconductor Digest.

For Jason Wang, a media operator based in Zhengzhou in central China’s Henan province, switching his long-held investment strategy of buying gold to tech-themed funds was a bittersweet decision.

The 28-year-old retail investor, who has no idea about tech company fundamentals, started buying tech funds – focused on China’s semiconductors, robotics and similar sectors – at the end of last year amid a bull run that lasted until the middle of this year. Watching returns on his semiconductor funds...

MIPS has introduced advances in its Actus reference platform, an open, energy-efficient RISC-V architecture designed to combine real-time motor control, edge inference, and communications for robotics.

Developed for compact robot joints, Actus addresses the requirements of humanoid and industrial robots, where precise movement must be achieved within strict space, weight, power, and thermal constraints. The [...] The post MIPS Introduces Actus RISC-V Platform for Robotic Motion Control appeared…

Texas Instruments Inc. (TI) has expanded its industrial communication portfolio with two devices designed to simplify connectivity across factory automation, robotics, and industrial IoT (IIoT) systems.

The TICS100MOD multiprotocol communications system-on-module (SoM) and TIOM2212 dual-channel IO-Link master provide hardware and software building blocks to connect field devices to industrial networks, reduce development complexity, and support [...] The post TI Launches Multiprotocol Industria…

The new Google TPU 8tx pairs AMD EPYC Turin hosts with confidential computing while TPU 8i Max doubles memory bandwidth for inference The post Google TPU 8tx Pairs AMD EPYC Turin Hosts with Confidential Computing while TPU 8i Max Doubles Memory Bandwidth appeared first on ServeTheHome.

By: Anil Godbole, Senior Product Planning & Marketing Manager, Intel and Daniel Byrne, Senior MTS, DCG, Intel AI agent sandboxes are commonly deployed as microVMs with several gigabytes of memory allocated to each, yet only a small fraction of that memory is touched between tool calls.

This access pattern makes sandboxes a natural fit […] The post Optimize Memory TCO with Intel Flat Memory Mode for Agentic AI on Future Xeon Processors first appeared on Compute Express Link.

WILMINGTON, Mass., Nov. 9, 2026 /PRNewswire/ -- Analog Devices, Inc. (NASDAQ: ADI) today announced that the Company's Chief Financial Officer, Richard Puccio, will discuss business topics and trends at the UBS Global Technology and AI Conference, located in Scottsdale, Arizona, on...

Third-quarter revenue of $9.5 million exceeds revenue generated during the entire first half of 2026 Acquired photonics and advanced packaging businesses contributed to commercial revenue growth while supporting QCi's quantum technology roadmap Ends quarter with approximately $1.3 billion...

// // Entegris updates its long-term revenue CAGR target to ~15%, driven by customers’ increasing materials intensity and contamination control requirements Disciplined execution and capital efficiency expected to support significantly higher profitability and cash flow Board authorizes $500 million share repurchase program BILLERICA, Mass.--(BUSINESS WIRE)-- E…

Stephen Neal is the board's lead independent director and head of corporate governance. Experts say how Nvidia handles the probe is crucial to retaining its reputation.

Unveiling IPC-1783-CO2E: A Common Language for Product Carbon Footprints in Electronics 1608095_alyssa… Nov 9, 2026 Save Your Seat Date Dec 2, 2026 | 9:00 - 10:00am CST Event Type Webinar Customers are increasingly asking electronics manufacturers to report the carbon footprint of their products.

Without a common approach, customers and suppliers face inconsistent requests, heavy burdens, different methods, and data that's hard to compare. IPC-1783-CO2E sets out a standard way to calculate and …

Jensen Huang is practically synonymous with Nvidia, the chipmaker he co-founded in 1993 and turned into one of the most valuable companies on the planet.

In the latest episode of "The Circuit" based on the transcript provided in the file "ben-jay (12).txt", hosts Ben Beheron and Jay Goldberg cover a wide spectrum of technology and semiconductor news, opening with a discussion on the unexpected shortage of Apple's premium polishing cloths, which Ben argues are uniquely necessary for cleaning new nano-textured touchscreen devices.

The core of the conversation analyzes a recent, massive sell-off in semiconductor stocks—dubbed the "semi-apocalypse"—…

Goldman CEO David Solomon says investor appetite for AI infrastructure remains strong despite a reset in AI stocks and political uncertainty around data centers.

Amazon Elastic Kubernetes Service (Amazon EKS) now supports Dynamic Resource Allocation (DRA) for NVIDIA GPUs and Elastic Fabric Adapter (EFA) network devices when using Amazon EKS Auto Mode and Karpenter.

You can now define the device attributes your Kubernetes workloads need with DRA, share GPU devices across Pods, and allocate topology-aligned GPU and EFA devices while Amazon EKS Auto Mode and Karpenter dynamically scale Amazon EC2 instances based on workload demand. With DRA, you define Res…