NVIDIA unveils Vera Rubin – AI supercomputing goes full-stack
NVIDIA Vera Rubin platform visual – a full-stack AI supercomputing system powering the next era of training and inference.
At CES 2026, NVIDIA introduced Vera Rubin, a next-generation AI supercomputing platform. Breaking from its previous launches centered around chips, Vera Rubin brought CPU, GPU, high-speed networking, memory, and security all together in one optimized system tailored for large-scale AI training and inference.
First announced with Vera Rubin, NVIDIA signals a decisive shift: AI performance is no longer only about faster chips but about owning the entire AI compute stack from silicon to systems. In all, the platform promises major efficiency gains for complex foundation models and agentic AI workloads.
Why It Matters:
AI leadership is concentrating among the companies that are able to provide end-to-end infrastructure. As training costs are increasing and models become more complex, full-stack platforms like Vera Rubin redefine who can compete at the frontier-and at what scale.
Lenovo enters the AI assistant race with Qira
The Lenovo Qira visual – a cross-device AI assistant designed for personalized, always-on experiences.
Lenovo has also unveiled Qira as their new AI assistant that is expected to support cross-device functionality. Qira is aimed at competing with Apple and Google because of its focus on personalization and contextual understanding.
Instead of marketing it as yet another chatbot assistant like Qira, it seems that Lenovo is branding it as an integral part of its hardware ecosystem, which is an indication that AI assistants are no longer going to remain an added feature.
Why It Matters:
The landscape of the AI assistant competition is gradually transitioning from the capability of the model itself towards distribution and integration with devices. Lenovo’s action also exemplifies the manner in which the hardware players are employing AI in order to strengthen the ecosystem effect and compete head-on with operating system-level AI assistants.
France awards military AI contract to Mistral AI
Mistral AI Defense Collaboration Visual – sovereign AI models for national security and military use.
France’s Armed Forces Ministry signed a framework agreement with Mistral AI on September 16, opening the way for the employment of its own models for AI in the defense and strategic sectors and for research purposes. The agreement highlights the importance of using French-controlled infrastructure for hosting AI systems.
The trend is a sign of increased concerns by governments about being reliant on third-party AI vendors, especially in the military and security sectors.
WHY IT MATTERS
“Sober AI” will become a high priority in terms of national security. “A case in point worth mentioning is that of the French administration’s contract for the development of ‘Mistral AI,’ which shows how nations demand autonomy in matters of core AI and do not want to be too dependent on the US or China in terms of these capabilities,” explains
Liked this issue? Get the next one free.
One email on Tuesday. Unsubscribe in one click.