At the VivaTech technology conference in Paris, NVIDIA CEO Jensen Huang outlined a substantial investment in Europes burgeoning artificial intelligence capabilities. Huang detailed plans for scaling AI infrastructure across the continent, powered by NVIDIAs Blackwell architecture, and encompassing the development of ‘agentic AI’ – autonomous, reasoning systems – and sovereign cloud deployments. This initiative aims to establish a new ‘intelligence infrastructure’ for Europe, fostering both independent AI development and a robust industrial base capable of leveraging the technology for economic growth and innovation.
A New Industrial Revolution
At the core of this transformation are systems such as the GB200 NVL72 – a high-performance GPU and NVIDIA’s most powerful AI platform – now in full production. These systems support applications ranging from sovereign model development to quantum computing, designed for complex reasoning and planning utilizing substantial computational resources for internal processing and optimization. NVIDIA’s manufacturing partners are currently producing 1,000 GB200 systems weekly, with further expansion anticipated, offering systems ranging from the compact NVIDIA DGX to larger configurations.
NVIDIA is actively establishing an industrial AI cloud, with the world’s first instance slated for construction in Germany. This initiative aims to provide European manufacturers with the necessary tools to simulate, automate, and optimize processes at scale, representing a significant step towards widespread industrial digitization. Ongoing collaborations with numerous companies across the continent focus on constructing digital twins leveraging the NVIDIA Omniverse platform, facilitating advanced modelling, analysis, and optimization within European industries.
The Industrial Cloud Goes Live
Huang detailed ongoing collaborations with numerous companies across the continent, focused on constructing digital twins leveraging the NVIDIA Omniverse platform. These virtual representations of physical systems will facilitate advanced modelling, analysis, and optimisation, driving efficiency and innovation within European industries. The presentation highlighted a future where robotics permeates all aspects of movement and transportation, with NVIDIA DRIVE, the company’s full-stack autonomous vehicle platform, now in production, accelerating the deployment of safe and intelligent transportation solutions.
Furthermore, a demonstration featuring Grek, a small robot developed in partnership with DeepMind and Disney, showcased Newton, a cutting-edge physics training engine for robotics, illustrating the potential for advanced robotic capabilities. Huang emphasized that the simulations displayed during the keynote were not animations, but rather accurate computer simulations reflecting the inherent beauty of the world and the elegance of mathematical principles, underscoring NVIDIA’s commitment to creating realistic and accurate digital environments for industrial applications.
The Next Wave
The advent of a new wave of artificial intelligence is characterised by exponential growth, extending beyond virtual applications into the physical realm with robotics and ‘agentic’ AI. NVIDIA posits that the necessary technology to instruct robots in manipulation and simulation is now available, representing a significant leap forward in capability. This expansion is fuelled by a dramatic increase in inference workloads; user numbers have risen from 8 million to 800 million in just two years.
To accommodate this demand, NVIDIA highlights the necessity for specialised computing architecture, introducing Blackwell as a ‘thinking machine’ specifically designed for reasoning and complex processing. Blackwell-powered systems will form the core of a new generation of data centres, termed ‘AI factories’. These facilities will be dedicated to generating ‘tokens’ – the fundamental units of modern intelligence – and will underpin the operation of increasingly sophisticated AI applications. This infrastructure is intended to support not only virtual AI but also the operation of physical robots, as exemplified by the demonstration with the robot Grek.
Exponential Inference Growth
The expansion of artificial intelligence is currently characterised by exponential growth in inference workloads. NVIDIA reports a tenfold increase in users employing inference – rising from 8 million to 800 million in just two years – necessitating a fundamental shift in computing architecture. This surge in demand is driving the development of specialised hardware designed not simply for calculation, but for ‘thinking’ and ‘reasoning’, exemplified by the Blackwell architecture.
Blackwell-powered systems are intended to form the core of a new generation of data centres, termed ‘AI factories’. These facilities will be dedicated to the generation of ‘tokens’ – the fundamental units of data used by modern AI models. This focus on token generation highlights a move towards large-scale, automated AI processing, enabling more complex and nuanced applications. The current wave of AI development, therefore, is not merely incremental, but exponential. This growth is fuelled by both physical robots and ‘information robots’ – or agents – requiring increasingly sophisticated computational resources. The combination of robotics and AI agents necessitates a computing infrastructure capable of handling both real-world manipulation and complex information processing, driving the need for specialised hardware and dedicated AI factories.
The Future of AI and Robotics
NVIDIA envisions a future where these AI factories provide the computational resources necessary for advanced robotics and agentic AI, creating a symbiotic relationship between virtual and physical intelligence. The company emphasizes that this future is being built in partnership with Europe, leveraging sovereign infrastructure and fostering innovation in the field of artificial intelligence.
More information
External Link: Click Here For More
