➀ IQE and Quinas Technology announced the completion of a £1.1m UltraRAM industrialisation project, developing scalable GaSb epitaxy for memory devices;

➁ ULTRARAM integrates non-volatility of flash storage with the speed and efficiency of DRAM, targeting AI, quantum computing, and space/defense applications;

➂ The project leveraged Innovate UK funding and IQE's industrial capabilities, advancing sovereign memory technology in the UK semiconductor sector.

➀ Fraunhofer IDMT-HSA and CARIAD SE are testing new acoustic sensors and measurement technology in vehicles to enhance automated driving by detecting external sounds like emergency sirens and enabling voice interaction;

➁ The systems underwent rigorous testing in Sweden under extreme weather conditions (snow, ice, heat, and rain) to evaluate reliability and usability;

➂ The KI4BoardNet project focuses on developing AI-driven vehicle power and data networks, integrating sensor data with real-time vehicle metrics for future automotive applications.

➀ Siemens introduced generative and agentic AI capabilities across its EDA portfolio, enhancing semiconductor and PCB design tools including Aprisa, Calibre, and Solido.

➁ Aprisa AI improves productivity by 10x, speeds tapeout time by 3x, and boosts PPA metrics by 10%, while Calibre Vision AI halves violation resolution time with intelligent clustering.

➂ The AI system offers open, customizable workflows with multi-LLM support, prioritizing data control for users, and is now available for early access across Siemens' EDA tools.

➀ Semiconductor manufacturing equipment industry is projected to grow at a 7% CAGR, reaching 11.1 million wafers per month (wpm) by 2028, driven by a 69% expansion in advanced process capacity (7nm and below) to 1.4 million wpm.

➁ 2nm-and-below capacity is expected to surge from 200k wpm in 2025 to 500k wpm by 2028, with equipment investment skyrocketing 120% CAGR to $43 billion, supporting mass production of 2nm chips by 2026 and 1.4nm by 2028.

➂ AI-driven demand for powerful computing (training, VR/AR, humanoid robots) is fueling investments across the semiconductor ecosystem, highlighting the industry's role in tech innovation and chip demand.

➀ Liquid Instruments introduced AI-powered Generative Instrumentation, enabling engineers to build custom test setups via natural language input;

➁ The system automates complex configurations, adapts to test requirements, and deploys real-time FPGA-based solutions without manual coding;

➂ Moku:Delta hardware platform offers 2 GHz bandwidth and multi-channel support, praised by researchers for accelerating advanced signal analysis workflows.

➀ The Q1 2025 AR/VR headset market grew 18.1% year-over-year, with Meta dominating (50.8% share) and emerging brands like XREAL, Viture (268.4% growth), and TCL driving competition;

➀ Shift toward Mixed Reality (MR) and Extended Reality (ER) is reshaping the market, with MR/ER shipments projected to reach 15.2 million and 8.6 million units by 2029, respectively;

➀ Despite a temporary 12% shipment decline in 2025, the market is expected to rebound with a 38.6% CAGR through 2029, with traditional leaders like Sony and Apple absent from top rankings due to inventory management.

➀ VeriSilicon launched ZSP5000 DSP series IPs based on fifth-generation architecture for computer vision and embedded AI;

➁ The series features scalable, energy-efficient design targeting edge intelligence devices;

➂ Combined with AI software ecosystem partnerships, it enhances efficiency in automotive/industrial edge computing.

➀ Researchers from TU Graz and partners developed methods to run AI models on IoT devices with ≤4KB memory, enabling localized AI processing for applications like precision positioning;

➁ Techniques include modular specialized models, subspace configurable networks (SCNs), quantization, and pruning to optimize efficiency while maintaining accuracy;

➂ Potential applications span industrial automation (drone navigation), automotive security (keyless entry verification), and smart home systems with extended battery life.

➀ Researchers from TU Graz, Pro2Future, and the University of St. Gallen developed methods to deploy AI models on resource-constrained IoT devices with as little as 4KB memory, enabling localized processing without external compute resources;

➁ The team used techniques like model partitioning, Subspace-Configurable Networks (SCNs), quantization, and pruning to balance model size and accuracy, achieving up to 7.8x faster image processing on IoT devices;

➂ Applications include industrial automation (e.g., drone/robot localization), smart home remote controls, and anti-spoofing for keyless car entry systems, demonstrating broad scalability across embedded systems.

➀ The University of Bremen collaborates with industry partners to develop an AI-driven modular edge computing system for real-time status monitoring of mobile machines in logistics and construction;

➁ The solution combines an AI cloud platform and transfer learning to minimize dependency on extensive field data, enhancing cost-effectiveness;

➂ Emphasizes data security through encryption and offers a user-friendly app to improve machine availability and operational efficiency.

➀ Universität Bremen collaborates with ANEDO GmbH and SEGNO Industrie Automation to develop a modular edge-computing system combined with AI for real-time condition monitoring of mobile machinery in ports and construction;

➁ The system employs a cloud-based platform and transfer learning to analyze data, predict failures, and reduce reliance on extensive field data, improving cost efficiency;

➂ Enhanced data security measures with encryption and a user-friendly monitoring app aim to prevent unauthorized access and support industrial digitization.

➀ ESA and Leonardo collaborate on Machine Learning-based CRPAs to counter GNSS interference from jamming/spoofing;

➁ Project focuses on reducing antenna size/weight using ML optimizations, targeting space-constrained applications like aircraft;

➂ Aims to achieve Technology Readiness Level 4 in two years, including lab testing and comparison with traditional antennas.

➀ A Columbia Journalism Review study reveals generative AI search engines like ChatGPT, Perplexity, and Gemini provided incorrect answers to over 60% of queries, with Grok 3 failing 94% of tests;

➁ Premium AI tools performed worse than free versions, often delivering overly confident yet inaccurate responses;

➂ These tools bypassed source citations, fabricated links, and disrupted traffic to original content providers, challenging digital content ecosystems.

❶ The article emphasizes the importance of transparency in AI decision-making, particularly in fields like medical diagnostics and recruitment, where understanding the rationale behind AI outputs is critical for trust and model improvement.

❷ It highlights two main focuses of Explainable AI (XAI): enhancing data/model quality for engineers and addressing ethical requirements to provide user-centric explanations, ensuring responsible AI deployment.

❸ The whitepaper advocates advancing XAI research, standardizing tools for large-scale models, integrating XAI into AI education, and encouraging corporate adoption to foster collaboration between human expertise and machine learning.

➀ BlackSky Technology is expanding its Gen-3 satellite constellation with AROS satellites to provide multispectral, large-area imagery for geospatial and digital twin applications;

➁ The AROS satellites feature dynamic broad-area monitoring, low-latency data delivery via optical inter-satellite links (OISL), and integration with AI analytics platform Spectra;

➂ Scheduled for launch in 2027, the satellites aim to enhance real-time tasking, country-scale mapping, and maritime surveillance capabilities.

➀ SK Hynix unveiled a UFS 4.1 storage solution using 321-layer 4D NAND flash, marking the industry's highest layer count;

➁ It delivers 4300MB/s read speed, 15%-40% faster random operations, 7% power efficiency improvement, and ultra-thin 0.85mm design for AI smartphones;

➂ The 1Tb TLC chip will expand to SSDs, strengthening SK Hynix's position as a full-stack AI memory provider.

➀ Micron Technology introduces the world's first LPDDR5X memory using 1γ node technology, achieving 10.7 Gbps speed and 20% power savings;

➁ The ultra-thin 0.61mm package enables slimmer designs for smartphones and foldable devices, supporting on-device AI tasks like real-time translation and image generation;

➂ Performance tests show 30-50% faster AI responses compared to previous generations, with potential applications in AI PCs and automotive systems.

➀ Ed meets water company lobbyists concerned about mandatory reporting of data centers' water usage under EU's CSRD and UK pressure;

➁ High-workload AI data centers consume water equivalent to towns of 50,000 people, sparking community resistance fears;

➂ Despite ethical concerns, Ed considers supporting reporting delays due to water firms' history of lucrative payouts to executives and shareholders.

➀ Intel plans to outsource its marketing operations to Accenture, leveraging AI for tasks like information processing and personalized communications, resulting in significant layoffs within its marketing division.

➁ The restructuring aims to streamline operations, reduce costs, and focus internal teams on strategic projects, though the company has not clarified how AI will enhance brand strength.

➂ Affected employees will be notified by July 11, with some retained to train Accenture contractors during the transition phase.