Core Insights - Ambiq Micro, Inc. has announced the Atomiq system-on-chip (SoC), which integrates a Neural Processing Unit (NPU) aimed at enabling real-time, always-on artificial intelligence at the edge, setting a new standard for energy efficiency in edge AI applications [2][3] Group 1: Product Features - Atomiq is the first SoC to utilize sub- and near-threshold voltage operation for AI acceleration, achieving over 200 GOPS of on-device AI performance, which supports complex workloads like computer vision and multilingual speech recognition [5] - The SoC features dynamic power scaling through SPOT-based ultra-wide range dynamic voltage and frequency scaling (DVFS), allowing operation at lower voltage and power, thus enhancing intelligence capabilities [5] - Ambiq's Helia AI platform, along with AI development kits and the modular neuralSPOT SDK, provides a comprehensive hardware-software stack that optimizes performance while reducing power consumption and development cycles [5] Group 2: Market Applications - The Atomiq SoC expands Ambiq's edge AI portfolio, enabling high-performance, battery-powered devices that were previously limited by power and thermal constraints, applicable in smart cameras, wearables, and more [6] - Ambiq's partnership with Bravechip aims to reduce smart ring costs by up to 85% and improve production yield by 20%, facilitating the development of next-generation wearables with advanced AI features [8] - The collaboration with Ronds has led to the deployment of over 400,000 intelligent sensors in heavy industries, providing continuous monitoring and predictive maintenance without relying on cloud connectivity [10] Group 3: Future Roadmap - Ambiq is committed to advancing AI applications across various sectors, including smart buildings, healthcare, and consumer electronics, with plans to unveil details on its next-generation 12nm SPOT platform in March 2026 [12][13] - The Atomiq platform is positioned to support demanding future applications, such as conversational AR glasses and autonomous industrial robots, showcasing Ambiq's leadership in ultra-low power edge AI [7][12]

Enabling always-on audio, vision, and reasoning for next-generation edge AI devices AUSTIN, Texas, Jan. 06, 2026 (GLOBE NEWSWIRE) -- Ambiq Micro, Inc. (“Ambiq”) (NYSE: AMBQ), a recognized leader in ultra-low-power semiconductor solutions for edge AI, today announced Atomiq, the highly anticipated system-on-chip (SoC) integrating a Neural Processing Unit (NPU) designed to enable real-time, always-on artificial intelligence at the edge. Built on Ambiq’s Subthreshold Power Optimized Technology (SPOT) platform, ...

Core Insights - Lantronix has launched the SmartEdge.ai Gateway and SmartSwitch.ai, which integrate AI inference, power, and connectivity for Edge AI applications [1][2][4] Product Overview - The SmartEdge.ai Gateway is 5G-ready and enables real-time video analytics with integrated 5G connectivity, while the SmartSwitch.ai aggregates inputs from multiple PoE switches and IP devices for centralized analytics [2][5] - The platform is powered by Qualcomm's QCS6490, ensuring high-quality video processing and edge inference in environments where latency and reliability are critical [2][4] Key Capabilities - Smart Video Insights: Real-time object detection, tracking, and vehicle monitoring [6] - Security Monitoring: Intrusion detection through integrated camera feeds [6] - Building Automation: Control of access points, lighting, shades, temperature, and occupancy [6] - Predictive Maintenance: Automated network analysis, event detection, and reporting [6] - High-Performance Aggregation: Features 10 1G/10G open SFP+ ports and two 1G/10G RJ45 ports with a 10G/25G SFP28 uplink [6] - Autonomous Network Management: Remote deployment, orchestration, and monitoring to reduce operating costs [6] Market Impact - The global video surveillance market is expected to grow from approximately $73.8 billion in 2024 to about $147.7 billion by 2030, indicating a CAGR of around 12% from 2025 to 2030 [5] - Lantronix's new platform positions the company at the forefront of the expanding market for AI-enabled surveillance and intelligent buildings [5][8] Strategic Advantages - Entry into high-growth segments requiring real-time analytics and powerful on-site compute [7] - New recurring revenue streams from multi-site deployments and device orchestration [7] - Strengthened partnerships within the AI ecosystem, particularly with Qualcomm Technologies [7][8] - Opportunities for margin expansion through higher-value compute and analytics products [8]

证券日报网讯1月5日，泰凌微在互动平台回答投资者提问时表示，公司推出了TL-EdgeAI平台，支持谷 歌LiteRT、TVM等主流开源边缘AI模型，并可转换和运行TensorFlow、PyTorch、JAX等框架模型，该 领域的进展敬请关注后续对外披露内容。 ...

TAIPEI, Jan. 5, 2026 /PRNewswire/ -- Advantech (TWSE: 2395), a global leader in Industrial AIoT, announced today the integration of Intel® Core™ Ultra Series 3 processors, built on Intel 18A, into its upcoming portfolio of industrial-grade motherboards and Edge AI computers. This strategic milestone strengthens the long-standing collaboration between Advantech and Intel, delivering breakthrough AI performance and enabling a new generation of intelligent, high-efficiency edge solutions. Continue Reading Spee ...

Core Insights - Lantronix Inc. has introduced a new Drone Reference Platform aimed at accelerating UAV design cycles and reducing integration risks for drone OEMs [1][4] - The platform is designed to support mission-critical applications and enables OEMs to deploy AI models in real-world flight applications [3][5] - The global drone sector is projected to reach $57.8 billion by 2030, indicating significant growth potential for companies in this market [5] Company Developments - The Drone Reference Platform is NDAA- and TAA-compliant, allowing OEMs to move from concept to flight-ready prototypes in weeks [1][4] - The platform features the Lantronix Open-Q™ 8550 series µSOM based on Qualcomm® Dragonwing™ QCS8550, optimized for airborne AI workloads [4][6] - Lantronix is expanding its role as a strategic supplier in the defense, intelligence, and commercial UAV markets, moving beyond connectivity into mission-critical compute and AI infrastructure [5][7] Industry Context - The increasing requirements for onboard compute in drone operations are positioning them as Edge AI systems, essential for real-time perception and autonomy [2] - The platform supports various applications, including intelligence, surveillance, reconnaissance (ISR), security monitoring, and industrial inspection [5][6] - Lantronix is well-positioned to capture the growing demand for secure, high-performance Edge AI platforms that support scalable drone programs [7]

Core Insights - Ceva, Inc. is expanding its AI ecosystem by integrating Sensory's TrulyHandsfree voice activation technology into its NeuPro-Nano NPU, aimed at enhancing voice-first user experiences in ultra-low power devices [1][4] - The collaboration allows for seamless, on-device keyword spotting in battery-constrained consumer devices, ensuring privacy and performance without reliance on cloud connectivity [1][3] Company Overview - Ceva, Inc. specializes in AI-driven products, providing a portfolio of silicon and software IP that enables devices to connect, sense, and infer, essential for intelligent edge applications [6][7] - Sensory, Inc. develops on-device AI technologies, powering over 2 billion devices globally, with innovations spanning various industries including automotive and consumer electronics [5] Product Features - NeuPro-Nano is Ceva's smallest and most power-efficient Embedded AI NPU, designed for AI inference in AIoT devices, now enhanced with Sensory's voice activation models for applications like TWS earbuds and smartwatches [2] - The integration of TrulyHandsfree technology into Ceva's NeuPro-Studio AI ecosystem allows developers to quickly prototype and deploy voice-enabled features optimized for NeuPro-Nano [3] Market Impact - The collaboration is expected to accelerate the development of AI-driven, voice-controlled devices, transforming consumer experiences in smart home, automotive, and wearable applications [4]

Core Insights - Ambarella, Inc. has launched the CV7 edge AI vision system-on-chip (SoC) optimized for various AI perception applications, including 8K consumer products, enterprise security cameras, robotics, and automotive designs [1][3] - The CV7 SoC features significant advancements in power efficiency, consuming 20% less power than its predecessor, the CV5, due to its 4nm process technology [3][5] - The CV7 integrates multiple functional blocks, enhancing performance and reducing form factors and costs compared to multi-chip solutions [4] Performance and Features - The CV7 boasts over 2.5 times the AI performance of the previous-generation CV5 SoC, supporting simultaneous processing of CNNs and transformer networks [5] - It includes advanced image signal processing capabilities, such as high dynamic range (HDR) and improved low-light performance, achieving quality down to 0.01 Lux [6] - The hardware-accelerated video encoding capabilities have doubled performance compared to the CV5, enabling high-resolution video streams [7][8] Market Position and Applications - The CV7 is positioned to enhance video analytics and image quality in next-generation consumer and enterprise security products [3] - It supports a wide range of applications, including AIoT, industrial automation, and advanced driver assistance systems (ADAS) [10] - The SoC samples are currently available, with demonstrations taking place at CES in Las Vegas [9]

Core Insights - Etron Technology will showcase its innovations at CES 2026 under the theme "MemorAiLink® Show Up," focusing on edge AI solutions across memory, intelligent vision sensing, high-speed connectivity, and privacy computing [3][13] Group 1: Product Innovations - The RPC® inside G120 Subsystem has won the 2025 Hsinchu Science Park Innovative Product Award, highlighting Etron's leadership in AI vision and semiconductor integration [5] - Etron has developed a new scalable DDR3 product line and an ASIC AI memory solution aimed at supporting SLM/VLM computing for edge AI applications [6] - The EJ732 series has achieved USB-IF certification as a USB PD 3.2 DRP controller, supporting up to 240W (48V/5A) Extended Power Range and enabling USB4 80Gbps high-speed data transfer [7] Group 2: Robotics and Automation - Etron, in collaboration with eYs3D Microelectronics, has launched the AMR01C/AMR01M Robotics Barebone Platforms, designed for rapid customization and integration in robotics applications [8] - A strategic partnership with Asia Optical aims to build domestic delivery and robotic base platforms, contributing to a local robotics ecosystem [9] Group 3: Privacy and AI Solutions - The DeCloakBrain™ Privacy AI Robotic Decision System has received CES 2026 Innovation Awards Honoree status, reinforcing Etron's position in privacy computing and autonomous decision-making [11] - The "DecloakVision" system, which combines Differential Privacy and Homomorphic Encryption, has won the 22nd National Innovation Award and is deployed in major medical centers in Taiwan [12] Group 4: Market Context - The global robotics market is projected to exceed $100 billion by 2025, indicating significant growth potential for robotics applications [8] - Etron's innovations are positioned to enhance system performance and reduce design complexity, aligning with the growing demand for smart automation in the edge AI and robotics market [3][13]

TAIPEI, Dec. 31, 2025 /PRNewswire/ -- Advantech is proud to introduce its latest suite of high-performance edge AI compute solutions powered by the Dragonwing™ IQ-X platform – the AOM-6731, AIMB-293, and SOM-6820. Built on this groundbreaking platform, these innovative products are engineered to meet the demanding requirements of modern industrial applications by delivering exceptional processing power, integrated AI acceleration with up to 45 TOPS of AI performance, and robust and lightning-fast 5G and Wi- ...