News from November 2026

Even silicon starts behaving like rubber on a microscopic level. As hybrid bonding moves toward pitches of 1 micrometer and below, tiny deformations become a major obstacle.

Research by Besi and Imec shows why mastering the process is just as important as building a highly accurate bonding machine.

Taipei, Nov. 3 (CNA) MediaTek Inc., Taiwan's largest integrated circuit designer, has generated solid returns for investors over the past 25 years, as the company has diversified its product portfolio to boost revenue and pay good dividends, MediaTek's CEO said Tuesday.

Europe is scrambling to develop infrastructure to support the rollout of AI amid high energy and land costs.

Deep learning (DL) models for real-time wearable sleep staging often rely on long-kernel temporal convolutions, a requirement that current accelerators struggle to support efficiently.

To overcome this, we present an FPGA-based CNN-BiLSTM accelerator that incorporates several algorithm-hardware co-design techniques to enable efficient edge inference. Central to this design is an FFT-based non-unit stride fast convolution, which performs accumulation in the frequency domain to enable a single IF…

This letter presents a mixed-precision CNN-LSTM accelerator for EEG-based sleep staging.

To efficiently execute long temporal convolutions, a serpentine dataflow propagates activations between neighboring processing elements, reducing memory accesses by 6.2 times and 2.25 times compared with output- and weight-stationary dataflows, respectively.

A hybrid quantization scheme applies runtime dynamic quantization to the LSTM and MLP layers and assigns mixed precision on a layer-by-layer basis, red…

The rapid scaling of large language models (LLMs) has unlocked remarkable capabilities across various domains. However, deploying LLMs on mobile devices remains challenging because of their high computational and memory requirements.

Low-bit quantization (e.g., 4-bit weights, 8-bit activations) can reduce model size while preserving accuracy, but its practical efficacy on mobile GPUs has been limited so far. In this paper, we identify that the performance gap stems primarily from frequent data …

The growing computational demands of deep learning (DL) have driven AI-optimized reconfigurable devices, such as the AMD Versal AI Engines (AIEs), the Intel Stratix 10 NX Tensor Blocks (TBs), and the AMD Ryzen AI NPUs.

Their heterogeneous resources and novel dataflow models make efficient exploitation nontrivial, and FPGAs lack native support for structured sparsity. This dissertation optimizes dense general matrix multiplication (GEMM), the dominant DL kernel, on these devices, proposes FPGA a…

Geometric perception on edge must simultaneously achieve two goals that are rarely addressed together: preserve accurate spatial understanding for object localization and operate within stringent latency, memory, energy, and precision constraints.

This coupling is particularly evident in oriented bird's-eye-view (BEV) detection, where global context, multi-scale interactions, and geometric fusion affect both prediction quality and hardware dataflow. Conventional deployment flows first compress …

Deep neural network (DNN) accelerators for edge inference must meet tight energy, latency, and area (EDP) constraints, and their optimal hardware architecture and dataflow configuration is typically identified through design space exploration (DSE).

However, on-chip memory technology including cell type, bank organization, and process node, is rarely treated as a first-class citizen in DSE. For example, increasing memory capacity can reduce memory accesses through greater data reuse, but at the…

Randomized caches protect against side-channel attacks by obfuscating the mapping of memory addresses to cache locations. Existing proposals typically describe architectural behavior, leaving practical hardware implementation challenges largely unexplored.

We implement four prominent designs, CEASER-S, ScatterCache, SassCache, and MIRAGE, and integrate them as the last-level cache (LLC) of a multi-core out-of-order RISC-V system-on-chip (SoC). We identify common and design-specific implementati…

Large language models (LLMs) are increasingly used to generate analog circuits, yet every netlist must also be verified. We propose LLMs as verification assistants for analog circuits that flag unmet specifications and reliability issues that appear under process, voltage, and temperature (PVT) variation.

These issues emerge from how circuit stages combine, a structure that generic parameter-efficient fine-tuning (PEFT) methods such as LoRA do not capture. To this end, we introduce CAPER, a cir…

Residual-stream caching can reconstruct transformer key-value (KV) caches losslessly, but its memory advantage depends on architecture. We present AnchorKV, a storage-replay spectrum connecting per-layer residual caching and full recomputation through configurable residual anchors.

An architecture-dependent break-even rule identifies the first storage-saving spacing, while deployment selects an operating point from the spectrum and cascade replay reduces reconstruction from O(Ls) to O(L), with …

The $115 billion deal aims to create a category leader in AI connectivity by co-designing network processors and lasers for optical transceivers. The post Marvell Snaps Photonics Chipmaker Lumentum to Bolster AI Connectivity appeared first on EE Times.

Wi-Fi, Bluetooth LE, Thread, and on-device AI on a single chip, with existing software and tools, built for billions of intelligent devices that run for years on a battery OSLO, Norway, Nov. 3, 2026 /PRNewswire/ -- Nordic Semiconductor, a global leader in low-power wireless solutions,...

MUNICH, Nov. 3, 2026 /PRNewswire/ -- electronica 2026 will take place November 10–13 in Munich, Germany. As AI moves from the cloud to edge devices, Longsys(301308.SZ/09976.HK) will showcase new products and edge AI storage solutions under the theme "Integrated Storage Empowering Edge...

Cities generate enormous amounts of data from cameras, sensors and operational systems, but turning it into useful action — for planning, maintenance and public services — remains a challenge.

Physical and agentic AI offer new ways to close that gap, from digital twins for infrastructure planning to vision and service agents, while keeping local governments […]

Blackstone President Jon Gray says AI infrastructure needs more equity as debt investors face concentration risks amid massive spending plans.

Infineon and Qualcomm are working together to accelerate the development and deployment of central car computers in software-defined vehicles (SDVs). The companies have signed a technology agreement to have a […] The post Infineon and Qualcomm hook up for in-car computing appeared first on Electronics Weekly.

Noak, a spin-off of imec and UHasselt, has raised €5 million to scale and industrialize the development of the next generation of solar: perovskite–silicon tandem PV. The seed investment, provided […] The post Imec spin-off Noak raises €5m appeared first on Electronics Weekly.

Greece has opened a draft national strategy for semiconductors and microelectronics for public consultation, setting out how the country hopes to build a larger role in chip design, research and […] The post Greece unveils semiconductor strategy appeared first on Electronics Weekly.

Marvell is to buy photonics specialist Lumentum for $115 billion in what is being seen as a bet on the continuation of the datacentre build out. “AI infrastructure is undergoing […] The post Marvell buys Lumentum appeared first on Electronics Weekly.

Accessing external memory puts the brakes on physical AI, so carmaker has found an economical and speedy shortcut

The European Union's Cyber Resilience Act (CRA), which becomes mandatory in December 2027, is raising the bar for cybersecurity in connected devices. Rather than relying on software patches to address vulnerabilities after they emerge, manufacturers are under growing pressure to build security into products from the outset.

For connected devices, that shift puts flash memory in a new role: alongside storing code and data, it must help protect device identity, data integrity and software updates…

World’s biggest chipmaker tells cloud computing providers a majority of their clients must be non-Chinese

Amazon Relational Database Service (Amazon RDS) for Oracle now supports M8i and R8i instances in additional AWS Regions. Both M8i and R8i instances are now available in Asia Pacific (New Zealand) and Canada West (Calgary), and R8i instances are now available in Europe (Milan, Stockholm, and Zurich) and Israel (Tel Aviv).

These instances are powered by custom Intel Xeon 6 processors, available only on AWS, delivering the highest performance and fastest memory bandwidth among comparable Intel pro…

Outlook examines how AI-driven demand is reshaping supply, technology investment, sustainability and the products that reach consumers. The post TechInsights Identifies Five Trends Reshaping the Semiconductor Market appeared first on Semiconductor Digest.

Researchers investigate how ultrashort laser pulses can be used to slice diamond, with potential applications in optics, semiconductors, and power electronics. The post Laser Slicing of Large-Area Diamond Substrates appeared first on Semiconductor Digest.

2026 Excellence in Achievement Award honors the company's leadership in expanding apprenticeships, investing in employee advancement, and strengthening the semiconductor workforce. The post SEMI Foundation Recognizes FM Industries for Advancing Semiconductor Workforce Development appeared first on Semiconductor Digest.

Marvell announced its intention to acquire Lumentum in a $115B acquisition to further optical connectivity in the AI era The post Marvell Proposes Acquiring Lumentum to Combine Optical Circuit Switching with AI Connectivity Silicon appeared first on ServeTheHome.

Building cloud solutions with autonomous AI agents should feel fast, safe, and reliable. Without cloud-specific context, coding agents often struggle.

They might invent deprecated SDK syntax, guess invalid command flags, or attempt operations without the required Identity and Access Management (IAM) permissions. Even worse, an unguided agent can enable billable APIs or provision cloud resources without your explicit permission.

Starting today, you can install the google-cloud-developer plugin d…

Applied Brain Research (ABR) has announced the general availability of its speech AI software development kit (SDK) for Zephyr RTOS on the Alif Semiconductor Ensemble family of edge AI processors.

The software enables streaming automatic speech recognition and text-to-speech synthesis to run entirely on-device, without requiring a cloud connection. It is designed for low-power embedded [...] The post ABR Launches Speech AI SDK for Alif Processors appeared first on Embedded.

Researchers at Materials Center Leoben, PVA TePla, and Advanced Semiconductor Engineering (ASE) published “AI-Based Image Enhancement Improves Detection of Interconnect Defects in Microelectronic Packaging.”

Abstract Excerpt: “Traditional failure-analysis (FA) methods face growing challenges due to the increasing miniaturization and complexity of modern semiconductor devices. In particular, the detection of cracks in the extreme low K... » read more The post AI Image Enhancement Improves Detect…

IP house believes optics demand is going to explode as Nvidia and others eye bigger systems

Intel’s turnaround now hinges on proving its foundry strategy can deliver, as customers look for signs the company will add fab capacity and catch TSMC

How far can GPU CFD scale? Convergent Science ran 536 million cells on two NVIDIA HGX B200 instances in 3.62 hours.

Here is how to size and scale your own simulations.

Sweta Gupta is the Vice President of Engineering at Analog Bits, where she leads development and execution of mixed-signal IP’s, managing circuit and layout teams.

Ms. Gupta has been a part of the Analog Bits team since 2004 and she has successfully managed and delivered high quality IP’s including low jitter PLL’s, low power crystal… Read More The post Interview with Sweta Gupta, VP of Engineering of Analog Bits appeared first on SemiWiki.

Researchers at KU Leuven and EURECOM published “SecDMP: Provably Secure Data Memory-Dependent Prefetching.” Abstract Excerpt: “In this work, we address this gap by designing a defense that is both secure and efficient.

First, we propose a classification of DMP leakage, categorize existing and novel attacks, and analyze which leakage classes are mitigated by existing defenses.... » read more The post Hardware Defense Blocks Memory-Dependent Prefetcher Leakage (KU Leuven) appeared first on Semico…

Researchers at IBM Research published “Meter-scale quantum interconnects between large quantum processors.” Abstract Excerpt: “Modular quantum architecture requires long-distance quantum interconnects to stitch together individual processor modules.

Such interconnects, long-distance couplers or L-couplers, must support coherent quantum operation over meter-scale distances and scale to multiple connections between any two modules. Here we demonstrate four L-couplers... » read more The post Quant…

Researchers at imec and Vrije Universiteit Brussels published “Embodied Emissions of 2.5D and 3D Scaling in Chipletized GPUs for Low-Carbon AI.” Abstract “GPU compute resources are scaling to support growing AI workloads, leading to further rises in the climate change impact of hyperscale data centers that rely on high-performance GPUs.

To model how packaging decisions... » read more The post 3D Integration Reduces Embodied-Emissions Impact Of GPU Scaling (imec, VUB) appeared first on Semicondu…

Tachyum files for bankruptcy, the court in Bratislava initiates bankruptcy proceedings.

Ben Bajarin and Jay Goldberg unpack a packed week for AI and semiconductors, starting with Anthropic’s S-1 and what it reveals about the cost of staying at the frontier.

We discuss why enterprise AI adoption still has room to grow, what our latest CIO survey says about ROI, and why Anthropic wants more control over its compute infrastructure. Hyperscaler earnings raise a broader question: how much of that AI investment is improving the rest of their businesses?

Then we turn to Intel’s disappoin…

Modern AI models are opening up new possibilities at the edge. Turning those possibilities into responsive, efficient applications, however, requires more than choosing hardware with an impressive TOPS figure.

Developers need to understand how models behave on their target devices, where performance is lost, and which optimisations… Read More The post WEBINAR: Porting, Analysing & Optimising AI Models onto Edge GPUs appeared first on SemiWiki.

Researchers at the University of Texas at Austin published “APPlace: Co-Optimizing Device Locations and Access Points for Routing-Aware Analog Placement.”

Abstract “Analog layout automation remains challenging in Analog/Mixed-Signal circuit design, because parasitics and mismatches introduced during layout strongly affect circuit performance.

Placement is especially critical among the layout stages, as it constrains the routing solution... » read more The post Co-Optimizing Access Points Improv…

Dive into the fundamentals of system technology co-optimization (STCO) and how it helps streamline chiplet-based designs.

Data infrastructure semiconductor solutions provider Marvell Technology Inc of Santa Clara, CA, USA has agreed to acquire Lumentum Holdings Inc of San Jose, CA, USA (which designs and makes optical and photonic products for optical networks and lasers) in an all-stock transaction that has been unanimously approved by the boards of directors of both companies...

US intelligence agencies reportedly concluded that a Chinese businesswoman with access to Nvidia CEO Jensen Huang was passing information to Beijing, raising questions about the security of the company’s commercial relationships and the enforcement of US semiconductor export controls.

The alleged Jensen Huang spy plot was reported by The New York Times on 29 October […] The post Alleged Jensen Huang spy plot highlights Nvidia export concerns appeared first on eeNews Europe.

I’ve been working with AMIQ EDA for almost nine years now, writing about their cool technologies and interviewing users about their experiences. I’ve read through more user guides and release notes than I can count.

I’ve watched many dozens of video clips showing how to use their products. I’ve even visited their headquarters … Read More The post Using for AI for Root Cause Analysis of Simulation Failures appeared first on SemiWiki.

Written by: Christopher Romano, Codi Starks, Garrett Mahoney, John Wolfram, Josh Murchie, K Singh, Mathew Potaczek, Mike Lombardi, Rik Dolfing Introduction Mandiant responded to multiple incidents in 2026 in which threat actors hijacked enterprise access to Large Language Models (LLMs), a tactic known as "LLMJacking."

Spanning a variety of cloud and LLM providers, these intrusions impacted organizations in North America, South America, and Asia Pacific across the financial services, retail, tec…

Imec and Belgium’s Ministry of Defense have announced a strategic partnership to accelerate the development of sensors, AI and secure connectivity for defense and security applications.

Cardiff University and Swansea University are to work together to explore the development of the UK's first multi-institution approach to industry-connected national compound semiconductor pilot foundry capabilities...