10+ hour, 29+ min ago   (464+ words) The value in datacentres is shifting from the chip to the rack, says Yole, with hyperscalers and AI labs co-designing their own AI ASICs and with advanced packaging, HBM4, co-packaged optics, and power (1 MW racks, 800 V, liquid cooling) setting the direction....

3+ day, 10+ hour ago   (207+ words) Infineon is sampling a dual-phase smart power stage family designed for AI accelerators and vertical power delivery modules. Integrating Infineon’s OptiMOS 6 MOSFETs and a dual-phase driver IC in a compact 6 x 6 x 0.8 mm³ package, the new family delivers benchmark power density exceeding…...

1+ week, 10+ hour ago   (151+ words) 2Q 2026 marked a major inflection point as switch sales in AI back-end networks surpassed those in front-end networks for the first time, reports Dell’Oro. Meanwhile, Ethernet further extended its lead in AI back-end networks, helping mitigate supply constraints. Ethernet AI back-end…...

2+ week, 6+ day ago   (178+ words) Starcloud, the space datacentre manufacturer and operator, announced a $250 million Series A extension funding round to its March $170m Series A, at a $2.3 billion valuation. Investors were Manhattan West, Benchmark, EQT, Soma, NFX, 776, and others, as well as new investors Nvidia,…...

2+ week, 6+ day ago   (139+ words) .SpaceXAI’s planned first-generation Starmind datacentre satellite, due for launch next year, will be based on the Nvidia Vera Rubin NVL72 rack-scale system, extending the Earth-bound Nvidia datacentre environment into space, announced Nvidia yesterday. SpaceXAI is the amalgamated SpaceX and xAI. Nvidia…...

3+ week, 10+ hour ago   (194+ words) HexSeed Technology, founded last year in Banchory near Aberdeen, has raised more than £600,000 in early-stage funding to scale its low-temperature diamond coating technology designed to prevent AI power semiconductors from overheating. Investors were Carbon13, Net Zero Technology Centre and Vento. Hexseed…...

3+ week, 6+ day ago   (219+ words) On 21 October, Electronics Weekly will be hosting the Energy Efficiency in the AI Age virtual event, bringing together engineers, innovators and technology leaders. Improving AI performance per Watt will be one of the key challenges that the event will cover....

4+ week, 10+ hour ago   (194+ words) The rise of inference workloads is driving new demand and introducing network requirements that differ from those associated with training workloads, says Dell’Oro. This shift is leading to significant expansion of front-end networks to support the broader buildout of AI…...