Advanced RF Systems, Agilex FPGAs, a Neuromorphic Chip for AI Optical Inspection: Embedded Week Insights

Here’s a roundup of this week’s must-read articles. We’ll cover the latest developments on advanced RF systems for mil-aero communications, Agilex FPGAs, and a neuromorphic chip for AI optical inspection. Also, check News Archives – Embedded and Technical Articles Archives – Embedded for the complete list of news and articles from our website. NEWS Aetina Introduces Palm-Sized In-Vehicle Edge AI Platforms The fanless systems combine Nvidia Jetson Orin, four GMSL2 camera inputs, and rugged connectivity for real-time AI vision at the edge. BrainChip, Orama.AOI Target AI-Based Industrial Defect Detection Optimized Akida models target high-accuracy inspection in tire production and semiconductor packaging while reducing edge AI power and system costs.

embedded.com

Vector Brings Natural-Language AI Automation to CANoe

Vector Informatik GmbH has expanded its CANoe development and testing environment with AI agents and Model Context Protocol (MCP) capabilities. The new functions allow users to describe development, analysis, and testing tasks using natural-language prompts, after which specialized AI agents can determine the required steps and execute complete workflows automatically. The CANoe AI Package and integrated MCP Server are available with CANoe 20 SP2. The approach can automate tasks ranging from generating individual code segments to orchestrating complete test procedures. For example, an AI agent can convert a software requirement into a CAPL test, execute it in CANoe, analyze failures, modify the CAPL code, and rerun the test.

embedded.com

Basler, Syslogic Deliver Industrial Machine Vision Solutions

Basler and Syslogic are expanding their partnership to enable machine vision and edge AI in industrial applications that require high-performance cameras and rugged AI computing platforms. The partnership leverages Basler’s ace 2 GMSL cameras and Syslogic’s rugged computers to deliver machine vision to applications such as autonomous mobile robots, agricultural vehicles, construction equipment, and intelligent transportation systems. Enabling the capture and analysis of surroundings in real time, the joint solution encompasses Basler’s ace 2 GMSL cameras, developed for embedded vision applications, and Syslogic’s rugged computers to create an optimized hardware platform for demanding vision AI applications, while reducing design complexity.

embedded.com

Advanced RF Systems Power Next-Gen Mil-Aero Comms

The radio-frequency (RF) industry, relying on established technologies for many years, is entering a significant phase of growth, driving the need for advanced solutions. Rising investment in defense and security systems and the expansion of the space economy are the two key drivers behind this growth. On the defense side, there is the demand for advanced RF technologies used in counter-drone systems, electronic warfare (EW), radar, and secure military communications. At the same time, large commercial and government satellite constellations are being deployed. These include low-Earth-orbit networks for broadband connectivity, Earth observation, navigation, and defense.

embedded.com

Aetina Introduces Palm-Sized In-Vehicle Edge AI Platforms

Aetina Corporation introduced the DeviceEdge AIE-VN34/44 and AIE-VO24/34, its first palm-sized in-vehicle edge AI systems. Based on the Nvidia Jetson Orin NX and Jetson Orin Nano modules, the platforms provide up to 100 TOPS and 67 TOPS of AI performance, respectively, and support four GMSL2 camera connections. The systems target smart transportation, autonomous mobile robots (AMRs), and industrial automation applications requiring local multi-camera processing in space-constrained environments. The AIE-VN34/44 systems use Jetson Orin NX modules with 8-GB or 16-GB memory and deliver up to 100 TOPS. The AIE-VO24/34 platforms use Jetson Orin Nano modules with Super Mode and 4-GB or 8-GB memory, providing up to 67 TOPS.

embedded.com

BrainChip, Orama.AOI Target AI-Based Industrial Defect Detection

BrainChip Inc. has partnered with Atlanta-based Orama.AOI to develop AI-based automated optical inspection solutions using BrainChip’s Akida neuromorphic AI platform. Orama.AOI has retrained and optimized Akida models using its industrial inspection datasets, demonstrating the platform’s application to production-oriented defect detection. Orama.AOI provides AI-powered automated optical inspection technology for industrial applications with current use cases including tire manufacturing and semiconductor packaging. By combining its proprietary inspection datasets with optimized Akida models, the company is developing edge AI solutions designed to identify defects with high accuracy while maintaining low power consumption. The collaboration demonstrates the role of specialized edge AI hardware in industrial…

embedded.com

SensiBel and Aizip Combine Optical MEMS and Edge AI Beamforming

SensiBel and Aizip are collaborating on a speech-processing solution that combines sensiBel’s optical MEMS microphone technology with Aizip’s edge AI beamforming algorithms. The companies aim to improve speech clarity and recognition in demanding acoustic environments, including smart glasses, true wireless stereo (TWS) earbuds, and industrial speech applications. Aizip’s beamforming solution combines AI-based noise reduction, dereverberation and directional speech enhancement. The processing is designed to improve the quality of speech signals delivered to automatic speech recognition (ASR) systems, particularly in noisy or far-field conditions. According to test results from the companies, Aizip’s processing can reduce the character error rate (CER) by more than 30%.

embedded.com

Humanoid Robots, Simplifying Embedded System Design, Photonic IC Development: Embedded Week Insights

Here’s a roundup of this week’s must-read articles. We'll look into the latest developments on humanoid robots, simplifying embedded system design, photonic IC development, and more. Also, check News Archives – Embedded and Technical Articles Archives – Embedded for the complete list of news and articles from our website. NEWS OpenLight and Tower Accelerate Photonic IC Development The expanded design ecosystem integrates photonic IP, semiconductor manufacturing, and established IC design workflows for optical applications. TI Unveils CAN XL Transceiver The CAN XL transceiver delivers real-time, high-bandwidth communication for industrial systems.

embedded.com

Altera Expands DDR5 and LPDDR5 Support Across Agilex FPGAs

Altera Corp. has expanded memory support across its Agilex FPGA portfolio with Quartus Prime Pro Edition Software version 26.1.1. The release adds DDR5, LPDDR5, and LPDDR5X-compatible memory capabilities to additional Agilex devices, giving developers greater flexibility when optimizing system bandwidth, power consumption, PCB space, and component availability. For Agilex 7 M-Series devices, the latest software release enables support for DDR5-6400 and LPDDR5-6400 memory. These interfaces provide higher memory bandwidth for demanding applications such as artificial intelligence, networking, and embedded computing. The new capabilities complement existing DDR5 and LPDDR5 support on Agilex 5 devices, as well as new LPDDR5 support on Agilex 3 devices.

embedded.com

Humanoid Robots Need a Real-Time, Distributed Nervous System

The emulation of the numerous types of movement produced by the human body in humanoid robots is arguably among the most complex challenges at the forefront of modern engineering. Designing robots to tackle everyday tasks such as assembly on a production line or cooking in a kitchen demands the coordination of multiple sub-tasks. The automation architecture to solve this challenge borrows as much from biological evolution as it does from the control theory that enables extremely fine motion control using electromechanical motors and actuators. Traditional robot design has focused on top-down motion policies that follow calculated trajectories.

embedded.com

OpenLight and Tower Accelerate Photonic IC Development

OpenLight and Tower Semiconductor Ltd. have expanded the design ecosystem for Tower’s PH18DA indium phosphide (InP)-on-silicon photonics platform. OpenLight’s photonic process design kit (PDK) is now available within Cadence Design Systems' electronic design automation (EDA) tools, enabling developers to design photonic integrated circuits (PICs) using an established IC design environment. The PDK has been developed through technical collaboration between OpenLight and Tower and combines OpenLight’s photonic intellectual property (IP) with Tower’s semiconductor process and manufacturing capabilities. Support for Cadence’s EDA environment is intended to simplify PIC development by bringing photonic design workflows closer to those already used for electronic ICs.

embedded.com

TI Unveils CAN XL Transceiver

Texas Instruments (TI) has claimed the industry’s first commercially-available controller area network (CAN) extended data-field length (XL) transceiver for industrial systems including humanoid robots, industrial robots, and HMI systems that require more data at faster speeds. The TCAN6062 CAN XL transceiver supports payloads up to 2,048 bytes per frame and data rates as high as 20 Mbits/s and ensures that priority messages are delivered first and on time. The TCAN6062 is backward compatible with CAN flexible data rate (FD) and CAN signal improvement capability (SIC), enabling a seamless migration in existing designs.

embedded.com

Microchip Upgrades PolarFire FPGA Ethernet Sensor Bridge

Microchip Technology Inc. has introduced Revision 2.0 PolarFire FPGA Ethernet Sensor Bridge for Ethernet-based AI inferencing at the edge. The smaller, production-ready board for standardized sensor integration, leveraging Nvidia Holoscan Sensor Bridge (HSB) technology, supports twice the number of cameras and offers a lower price point than the first generation. Thanks to the power efficiency, security, and reliability of PolarFire FPGAs, the second-generation HSB-enabled board offers real‑time, multi‑sensor data processing over Ethernet in power‑and space-constrained edge environments. It includes built‑in security and safety features to help protect edge AI devices, and integration with the 

embedded.com

Nordic nRF9151 Enables Cellular and Satellite IoT Devices

Nordic Semiconductor is simplifying the development of IoT devices that require connectivity across both terrestrial cellular and satellite networks. LooUQ’s MTC2-N9151 embedded modem has received Skylo non-terrestrial network (NTN) certification, building on the certification of Nordic’s nRF9151 cellular IoT module. This enables developers to use a pre-certified hardware foundation for products that can switch between cellular and satellite connectivity. The nRF9151 combines an application MCU, LTE-M/NB-IoT or DECT NR+ modem, NB-NTN satellite connectivity, and security functions in a compact module. With firmware supporting 3GPP NB-NTN, the module can communicate through Skylo’s standards-based satellite network when terrestrial cellular coverage is unavailable.

embedded.com

Simplifying Embedded System Design with a Modern Software Ecosystem

ADI details how an open, modular software ecosystem, using its CodeFusion Studio, can reduce development effort in embedded system design.

embedded.com

Silicon Motion Unveils PerformaShape Technology for AI SSDs

Silicon Motion Technology Corp. has introduced its MonTitan SSD reference design kit (RDK), incorporating the company’s next-generation PerformaShape technology. Presented at FMS2026, the platform is designed for agentic AI infrastructure, where enterprise SSDs are increasingly required to operate as a persistent memory tier for applications such as key-value (KV) cache offload. The solution targets predictable quality of service (QoS), sustained performance, and endurance under continuously changing AI workloads. Agentic AI applications place different demands on enterprise storage compared with conventional AI workloads. AI agents continuously process information, retain context, execute multiple inference steps, and interact with external tools.

embedded.com

HBM, Agentic AI Platform for PCB Design, Ku-Band RFoF Transmitter: Embedded Week Insights

Here’s a roundup of this week’s must-read articles. We'll look into the latest developments in high-bandwidth memory, Cadence’s agentic AI platform for PCB Design, and a Ku-Band RFoF transmitter. Also, check News Archives – Embedded and Technical Articles Archives – Embedded for the complete list of news and articles from our website. NEWS Partnership Delivers AI-Powered Design to Engineers A collaboration between Celus, Variscite, and NXP accelerates embedded system design. Samsung Advances AI Memory with New 3D Architecture New AI memory architectures, 400+ layer V-NAND, and next-generation HBM technologies target future AI systems.

embedded.com

Marvell Introduces New AI Memory Infrastructure Solutions

Marvell Technology has expanded its AI memory infrastructure portfolio with new solutions designed to address the growing memory requirements of large-scale agentic AI inference. The latest additions strengthen the company's offerings across AI storage, CXL-based memory expansion and pooling, and optical shared-memory architectures, enabling hyperscale cloud providers to scale memory resources more independently from compute. As AI models continue to increase in size and context length, memory bandwidth, capacity, and interconnect performance have become key factors influencing inference efficiency. Growing key-value (KV) cache requirements place significant pressure on conventional server-attached…

embedded.com

Last-Level Caches Matter Even More in the HBM Era

HBM dramatically increases bandwidth, but latency, power, and data locality still matter, so last-level caches have become more, not less, important.

embedded.com

Partnership Delivers AI-Powered Design to Engineers

Celus, an AI-powered electronic design automation company, announced a strategic partnership with Variscite that brings Variscite’s system-on-module (SoM) portfolio into the CELUS Design Platform to accelerate embedded system development from concept to a prototype or production-ready design. Simplifying electronics design with AI, the integration combines Variscite’s SoMs with Celus’ AI Design Assistant to shorten time to market through an automated process. Building on Variscite’s existing relationship with NXP Semiconductors, Variscite’s SoMs are now featured directly inside the NXP Block Diagram Designer, powered by Celus. This enables engineers to develop complex designs from start to finish in a single environment.

embedded.com

Samsung Advances AI Memory with New 3D Architecture

Samsung Electronics presented its latest memory technologies and long-term roadmap for AI infrastructure at the Future of Memory and Storage (FMS) 2026 conference, introducing new 3D memory concepts with advances in high-bandwidth memory (HBM), NAND flash, and enterprise storage. The announcements reflect the company's strategy of improving memory bandwidth, density, and energy efficiency to support increasingly demanding AI and high-performance computing (HPC) workloads. A key highlight of the event was the introduction of the concept models for zHBM and zNAND-O. The proposed zHBM technology differs from conventional HBM implementations by vertically stacking high-bandwidth memory directly above AI accelerators.

embedded.com

Silicon Labs Delivers its Lowest-Power Bluetooth LE SoC

The BG2B Bluetooth LE wireless SoC delivers high power efficiency, security, and integration.

embedded.com

Ambiq Launches heliaPROFILER for Edge AI Development

Ambiq has expanded its HELIA AI software ecosystem with the introduction of heliaPROFILER, an open-source profiling tool designed to accelerate the development and optimization of edge AI applications running on Apollo system-on-chips (SoCs). The software provides developers with cycle-accurate insight into AI model execution, enabling more efficient optimization of performance, memory usage, and energy consumption on resource-constrained embedded platforms. As edge AI models become increasingly complex, developers require detailed visibility into runtime behavior to balance computational performance with strict power and memory constraints. Traditional profiling methods often rely on multiple disconnected tools and manual data analysis, increasing development time and making optimization more difficult.

embedded.com

Cadence Develops Agentic AI Platform for PCB Design

The AuraStack AI super agent coordinates advanced PCB and package design with simulation and sign-off but leaves application-specific tradeoffs to engineers.

embedded.com

Optical Zonu Launches Miniature Ku-Band RFoF Transmitter

Optical Zonu has introduced a miniature Ku-band RF-over-fiber (RFoF) transmitter designed for defense and aerospace platforms where installation space is limited and high-frequency RF signal integrity is critical. The compact OEM module enables direct conversion of RF signals to optical signals at the antenna, reducing transmission losses and simplifying system architectures compared with conventional coaxial cabling. Measuring approximately 2.25 × 1.0 inches, the transmitter is intended for integration into phased-array antennas, satellite communication (SATCOM) terminals, radar systems, and other densely packed RF assemblies. By performing the RF-to-optical conversion close to the antenna, the module allows optical fiber to replace bulky coaxial cables, minimizing signal attenuation while providing greater flexibility for system designers.

embedded.com

Fault-Tolerant Quantum Computers, Sustained Performance Defines the Next Phase of Edge AI, Agentic AI for EDA: Embedded Week Insights

Here’s a roundup of this week’s must-read articles. We'll look into the latest developments on fault-tolerant quantum computers, why sustained performance will define the next phase of edge AI, and a partnership on agentic AI for EDA to automate complex engineering tasks. Also, check News Archives – Embedded and Technical Articles Archives – Embedded for the complete list of news and articles from our website. NEWS Renesas Upgrades DDR5 MRDIMM Chipset Gen 3 DDR5 MRDIMM chipset delivers speeds up to 16,000 MT/s. TDK, LG Innotek Collaborate on Physical AI Solutions…

embedded.com

Mikroe Integrates Swift Navigation Skylark with GNSS 19 Click

Mikroe has partnered with Swift Navigation to integrate the Skylark Precise Positioning Service into its GNSS 19 Click development board, providing users with a free six-month license to the cloud-based correction service. The collaboration is intended to simplify access to high-accuracy global navigation satellite system (GNSS) positioning for developers working on industrial, automotive, and IoT applications. The GNSS 19 Click is based on the Telit Cinterion SE868K5-D dual-frequency GNSS module, which supports simultaneous reception from multiple satellite constellations and frequency bands to improve positioning accuracy and reliability. By combining the onboard receiver with Swift Navigation's Skylark correction service, the platform can deliver centimeter-level positioning for applications that require precise navigation and tracking.

embedded.com

Renesas Upgrades DDR5 MRDIMM Chipset

Renesas Electronics Corp. has launched its third‑generation (Gen 3) DDR5 multiplexed rank dual in‑line memory module (MRDIMM) chipset solutions with server-class MRDIMM speeds up to 16,000 mega transfers per second (MT/s). The new devices include the RRG5013 multiplexed registering clock driver (MRCD) and RRG5103 multiplexed data buffer (MDB). Targeting increased demand for higher memory bandwidth driven by AI data centers, cloud infrastructure, and accelerated compute workloads, the Gen 3 MRDIMM solutions deliver 25% higher memory bandwidth than its second-generation solutions, while continuing to use existing DDR5 infrastructure. The MRDIMM platform includes MRCD, MDB, power management ICs, serial presence detect hubs, and temperature sensors.

embedded.com

Why Sustained Performance Defines the Next Phase of Edge AI

Avnet Silica details why sustained performance, not peak compute, will define the next phase of edge AI.

embedded.com

BrainChip Launches AKD1500 M.2 Module for Fanless Edge AI

BrainChip Inc. has introduced a new M.2 version of its AKD1500 edge AI accelerator, enabling compact, fanless, and plug-and-play deployment of neuromorphic artificial intelligence capabilities in industrial and commercial systems. Based on the Akida neuromorphic processor architecture, the module provides hardware designers with a low-power solution for adding on-device AI functionality to existing platforms without requiring major power or thermal redesigns. The AKD1500 M.2 module is designed as a drop-in AI acceleration solution for legacy systems that need enhanced processing capabilities while maintaining strict power and space constraints. By operating without active cooling, the module supports both portable and stationary designs where conventional AI accelerators may introduce excessive heat, power consumption or system complexity.

embedded.com