Microsoft plans to scale its Maia accelerators to multi-gigawatt deployments from 2027 and make them the main inference system for its own products. The Maia 300 runs scale-up and scale-out traffic over a single Ethernet fabric, but practitioners are questioning the tokens-per-watt figure at the heart of the announcement.
In an InfoQ talk, Andrew Martin argues that securing AI workloads is inseparable from securing Kubernetes. He covers unpredictable agent behaviour, GPU supply chain weaknesses and shared compute risks, along with defences ranging from prompt firewalls to identity for agents.
Volvo will develop an autonomous driving system comparable to Tesla's FSD using Nvidia's technologies and AI, including Alpamayo. The move marks a step up in the carmaker's self-driving ambitions.
The New Stack tested local AI models on the Surface Laptop Ultra, which Microsoft is positioning as an alternative to Apple's MacBook Pro for creatives and developers.
French cloud provider Scaleway plans two new AI computing sites in Sweden and Denmark. Both will combine AMD and Nvidia GPUs.
Qualcomm CEO Cristiano Amon said on Alex Heath's Sources podcast that leading AI companies are asking for phones able to run models of at least 100 billion parameters by 2028. The remark signals how far on-device AI ambitions are stretching.