In a fireside chat with Aude Durand, Deputy CEO of the epic Group, on Friday, Jensen Huang, the founder and CEO of NVIDIA, mentioned that the upcoming AI infrastructure would greatly benefit European startups and is timely.
During Scaleway’s ai-PULSE conference in Paris, Huang virtually addressed an audience of over 1,000 in-person attendees, stating, “We are currently witnessing significant upcoming waves.”
Speaking from Station F, the world’s largest business school, where Huang connected via video conference from NVIDIA’s Silicon Valley headquarters, he elaborated on two key driving factors behind this momentum.
Firstly, Huang emphasized “the recognition that every region and country must develop their unique AI capabilities.” Additionally, as AI concepts proliferate globally, Huang highlighted the “adoption of AI across various sectors.”
Huang pointed out the potential opportunities for Europe, especially in its thriving digital health and medical industries. He expressed his anticipation for advancements in fields like biotechnology, manufacturing, and robotics akin to those seen in speech technology, noting the presence of major commercial production companies in Europe.
Acknowledging France’s AI Leadership
Durand initiated the discussion by inquiring about Huang’s perspective on the AI landscape in Germany, particularly in France, where the government has invested millions in AI research and development.
Huang mentioned NVIDIA’s collaborations with 4,000 companies in Europe, with over 400 located solely in France. He highlighted partnerships with companies like Mistral, Qubit Pharmaceuticals, and Poolside AI, emphasizing Europe’s historical expertise in Artificial Intelligence.
Furthermore, Huang stressed the importance of establishing a robust technological ecosystem, citing Scaleway as a key player in fostering AI development not only in France but across Europe.
Huang emphasized the growing recognition among businesses of the significance of training AI models with region-specific data to account for the unique social and professional characteristics of each locality. This approach is gaining traction in Europe and beyond.
Scaleway and NVIDIA: Spearheading the AI Revolution in Europe
By offering sky credits to access their AI computing cluster featuring 1,016 NVIDIA H100 Tensor Core GPUs, Scaleway, a division of the prominent Western telecoms firm Iliad Group, is instrumental in driving the next wave of AI innovation in Europe.
Scaleway also provides a sovereign system ensuring compliance with EU data protection regulations, a critical aspect for companies operating within Europe.
Moreover, NVIDIA AI Venture software is accessible for download on the Scaleway Marketplace, catering to local participants of the NVIDIA Inception program, which supports startup development.
The program includes tools like the NVIDIA Triton Inference Server, NVIDIA TensorRT-LLM for enhanced inference, and the NeMo framework and convolutional models for various applications.
Leveraging Supercomputing Capabilities to Transform AI
Huang outlined NVIDIA’s rapid advancements in high-performance computing and AI on a global scale to empower the upcoming wave of AI applications.
These cutting-edge methodologies essentially function as “supercomputers,” with AI technologies currently among the most potent worldwide.
One notable advancement is Scaleway’s Nabuchodonosor computer, dubbed “Nabu,” equipped with 127 NVIDIA DGX H100 systems, aimed at assisting businesses in France and other European regions in scaling up AI initiatives.
Huang emphasized the unique capabilities of the Scaleway infrastructure, highlighting its exceptional computing power tailored for diverse AI tasks.
NVIDIA recently introduced the AI Foundry services on Microsoft Azure to expedite the development of innovative AI applications, catering to a broad spectrum of users, including renowned enterprises like Amdocs, Getty Images, SAP, and ServiceNow.
Huang also discussed the upcoming Venus supercomputer, located at the Jülich center in Germany, set to be Europe’s premier exascale AI system, boasting unparalleled computing power for a wide array of AI applications and simulations.
He elaborated on the HGX H200 AI processing system from NVIDIA, based on the Hopper architecture, featuring the L200 Tensor Core GPU, slated for release in Q2 of 2024 to redefine industry standards.
Furthermore, Huang detailed NVIDIA’s initiative to establish “AI factories,” advanced data centers powering diverse applications across sectors such as robotics, AI services, and electric vehicles.
Embracing Open Source
Addressing the role of open-source technology and available resources in AI, Huang expressed his strong support for open-source software, attributing much of AI’s progress over the past decade to this collaborative approach.
Huang emphasized the pivotal role of open source in fostering innovation, attracting researchers, and engaging businesses and scholars, underscoring its vital contribution to the ongoing advancement of AI technologies.
