The support or use of AI (artificial intelligence) in electronics, including ML (machine learning), whether in software (supervised, unsupervised or reinforcement learning tools) or hardware (accelerators, GPUs, etc).
A multi-domain passive module with functionality embedded in the substrate has been developed by Saras Micro Devices. The Saras Tile, or STILE, enables power regulation from system board to package. Saras’ approach is for vertical power delivery to reduce parasitic losses. Stacking the power stage and passive components in line with the die shortens the power path and reduces parasitic ...
Axelera AI, the Dutch provider of purpose-built AI hardware acceleration technology, has announced Europa – an AI processor unit (AIPU) that claims to set a new performance/price standard for multi-user generative AI and computer vision applications. Europa’s combination of processing power, energy and thermal efficiency, compact packaging, and multiple form factor options make it suitable for compute-intensive, multi-modal AI inference ...
The Ultra ECDP (electrochemical deplating) tool is designed specifically for wide bandgap compound semiconductor manufacturing, said ACM Research. It delivers uniform processing outside of the wafer pattern area with minimal undercut for high-precision wafer level gold (Au) bump, thin film and deep hole deplating. Gold exhibits high conductivity, corrosion resistance and malleability and is emerging as an advantageous material for ...
AI-powered intelligent power management integrated circuits are transforming how devices manage energy, says Samudrapom Dam. Power management integrated circuits (PMICs) use machine learning (ML) to predict load patterns and adjust voltage in real time to transform how devices manage energy. These systems extend battery life and address challenges such as compact integration, overheating and battery ageing. Edge-trained controllers and efficient ...
Researchers at CEA-Leti are redesigning the foundations of computing to meet performance needs and preserve the planet’s energy reserves. By Pat Brans. Artificial intelligence (AI) is now central to computing, but its appetite for power is growing unsustainably. As large language models, autonomous systems and edge applications proliferate, AI is demanding more from hardware than any previous class of workloads. ...
The 2026 IEEE/JSAP Symposium on VLSI Technology & Circuits has announced a call for papers around the theme: “Advancing the AI Frontier through VLSI Innovation.” The Symposium will be held from June 14 – 18, 2026 at the Hilton Hawaiian Village in Honolulu, HI (pictured). The Symposium paper submission site will open on December 2, 2025, and the deadline for ...
The ACS RTDI (Advanced Cloud Solution RealTime Data Infrastructure) software uses AI to test advanced chips, says Advantest. It uses low-latency edge computing and analytics to monitor complex semiconductor. The algorithms can provide analysis, corrective analysis and prompt actions to improve quality and yield. Advantest said, semiconductor manufacturers can use AI algorithms and data analytics to extract insights from the ...
A hyper-beam array lithography system has been introduced to the US, based on architecture pioneered in Europe. The lithography tool operates with 65,000 independently controlled parallel beams which delivers capabilities at a scale no electron beam technology can match, said the company. Installed at Northrup Grumann, it addresses the holy grail of chip design, said founder and CEO, Lex Keen. ...
The current supply of agentic AI models, platforms and products far exceeds demand, which will result in consolidation and market correction, according to Gartner. Agentic AI markets will consolidate in the short term as hype and fear of missing out (FOMO) give way to fundamental economics. In this, the losers of consolidation will be undifferentiated AI companies and their investors. ...
At Semicon West, Applied Materials unveiled three semiconductor manufacturing systems that it said targeted specific performance areas to meet AI computing demands. Kinex is an integrated die-to-wafer hybrid binding system, Xtera (pictured) is for GAA (gate-all-around) transistors and PROVision 10 eBeam is a metrology system for 3D chips. The Kinex bonding system was developed in collaboration with Dutch equipment supplier, ...
