AI

Northwestern University engineers have developed an artificial intelligence algorithm for smart robots that gather their own raw data. Dubbed ‘MaxDiff RL’ (maximum diffusion reinforcement learning), “the algorithm’s success lies in its ability to encourage robots to explore their environments as randomly as possible in order to gain a diverse set of experiences”, according to the university. “The algorithm works so ...

Leonardo will ready the UK’s Royal Air Force to handle future combat air mission data following the award of a new 10-year contract. Specifically, the Ministry of Defence (MOD)’s Defence Equipment and Support (DE&S) agency chose the company to develop and future proof the tools that generate and analyse the data used by the UK Armed Forces’ combat aircraft – ...

Blaize, the El Dorado Hills AI chip developer,  has raised $106 million from existing investors, including Bess Ventures, Franklin Templeton, DENSO, Mercedes Benz, and Temasek and new investors Rizvi Traverse, Ava Investors and BurTech LP LLC. The company had previously raised $224 million. Blaize’s programable  processor architecture and low-code/no-code software platform solutions enable enterprises to deliver AI capabilities at both ...

Japanese telco NTT has used machine learning to develop AI which predicts damage to outdoor communications infrastructure during disasters such as earthquakes or deluges. Resolution is as fine as individual telegraph poles. “This technology allows for accurate disaster prediction by building a model that learns disaster patterns from facility data that the NTT Group has inspected and repaired [after] past ...

Edge Impulse has introduced AI software for Arm microcontrollers and Nvidia processors intended to spot previously unseen and untrained anomalies in images, for industrial inspection, medical imaging and logistics, for example. It is based on ‘Gaussian mixture models’ – GMMs. “Neural networks are powerful but have a major drawback: handling unseen data, like defects in a product during manufacturing, is ...

In collaboration with Meta, NVIDIA has announced the availability of accelerated versions of Llama 3, optimized for NVIDIA GPUs, across cloud, data center, edge, and PC environments. Developers can access Llama 3 at ai.nvidia.com, where it’s offered as an NVIDIA NIM microservice with a standard API for deployment flexibility. Meta revealed its engineers trained Llama 3 on a computer cluster ...

Datacentre GPU and AI ASIC revenue could reach $156 billion by 2025 and $233 billion by 2029, says Yole Developpement. The massive growth that the data center GPU and AI ASIC market experienced in 2023 is expected to continue in 2024 before stabilising in 2025. Generative AI’s rise since late 2022 has boosted demand for high-speed memory technologies, including DDR5 ...

Rivos has raised $250 million to make datacentre AI processors using RISC-V cores. Rivos was founded by its CEO Puneet Kumar (pictured) who worked at P.A. Semi and was sued for theft of trade secrets by Apple when he left P.A. to found Rivos in 2021. “The rapid changes in LLMs and the merger with the data analytics stack makes ...

Supermicro has announced edge computers built around Intel’s Atom x7000RE processors. SYS-E100-14AM and SYS-E102-14AM mini-PCs have up to eight ‘Efficient’ cores had dissipate less than 12W. “While the SYS-E102 comes in a form factor of only 190 x 44 x 120 mm, these systems offer up to 16GB of DDR5-SODIMM, dual 2.5 GbE LAN ports, two independent HDMI ports, three ...

Mouser reflected one of the major themes at Embedded World 2024, intelligence at the edge, and announced it has signed Edge Impulse, the software framework provider for embedded machine learning (ML) applications. Machine learning is implemented at the edge of the network, explained Mouser’s Mark Patrick in order to make decisions very quickly at the [network] edge. [Edge Impulse’s] products ...