AI

Nvidia will bring out  a new family of AI chips every year, says CEO Jensen Huang. The next one, called Rubin, containing networking  ICs, a GPU and a CPU (called Versa) will be out in 2026. ASRock Rack, ASUS, GIGABYTE, Ingrasys, Inventec, Pegatron, QCT, Supermicro, Wistron and Wiwynn will deliver cloud, on-premises, embedded and edge AI systems using Nvidia GPUs ...

Ubotica, the Irish space AI specialist, has been selected by the European Commission to participate in its ARCHYTAS project, a three-year initiative to develop AI accelerators for defence applications. It means the Dublin-based company has been awarded an undisclosed grant that will be funded by the European Defence Fund (EDF). “We are delighted to receive our first grant from the ...

EdgeCortix, the five year-old Tokyo fabless startup specialising in AI-at-the-edge,  has launched its SAKURA-II Edge AI accelerator. Paired with EdgeCortix’s Dynamic Neural Accelerator (DNA) architecture, SAKURA-II is engineered to tackle Generative AI tasks. Designed for flexibility and power efficiency, SAKURA-II manages  tasks including Large Language Models (LLMs), Large Vision Models (LVMs), and multi-modal transformer-based applications at the edge. “SAKURA-II’s impressive ...

Dell  and NVIDIA have added infrastructure to the Dell AI Factory to service customers wanting ti  implement AI. “Organizations are moving quickly to capture the AI opportunity, which is why our collaboration with NVIDIA is so important,” says Michael Dell, “our expansion of the Dell AI Factory with NVIDIA continues our joint mission – we’re making it easy for organizations ...

Ambarella has announced a pair of 5nm AI camera processors for  in-vehicle fleet telematics. “CV75AX provides up to 2x the AI performance over Ambarella’s prior-generation SoC, enabling the latest transformer neural networks for improved accuracy and reduced false positives without having to train for every object. [It] also provides the AI horsepower for more complex driver-scoring models,” according to the ...

Rapidus, the Japan government-backed startup focussed on 2nm process technology, is to collaborate with RISC-V specialist Esperanto Technologies on developing AI chips. Esperanto develops  chip designs for AI/machine learning based on RISC-V. It is headed up by Dave Ditzel (pictured) who was the founder and CEO of the x86 clone startup Transmeta. Rapidus also has a RISC-V-based chip development agreement ...

Artemy is a fully-automated tomato harvesting robot, according to its maker Certhon Build. “It uses AI to determine the ripeness of cherry truss tomatoes,” said the company. “Only ripe tomatoes are selected, and cherry truss tomato peduncles are cut using scissors attached to the top of the harvesting robot arm and stored in the loaded crates.” Harvesting continues under direct ...

Researchers at Sandia National Laboratories aim to automatically generate code in Kokkos, a parallel programming language designed for use across many of the world’s largest supercomputers. The specialised language, developed by researchers from several national labs, handles the nuances of running tasks across tens of thousands of processors. Sandia is employing retrieval-augmented generation (RAG) to create and link a Kokkos ...

SoftBank and Arm are to develop proprietary AI chips and sell them as merchant market devices, reports the Nikkei. Softbank and Arm will  contribute to the development  costs and Arm will set up a separate division for the initiative. The chips will be fabbed  by foundries and TSMC is already said to have been approached for capacity. The plan is ...

Expanded to occupy six halls in the Nuremberg Messe, this year’s Embedded World was large and busy; it was impossible to see everything in three days, but here are some highlights. By Caroline Hayes. This year’s Embedded World welcomed more than 32,000 visitors, from more than 80 countries. This figure is nearly 20% more than 2023’s 27,000 visitors and the ...