Jensen Huang’s Vision for Europe Runs on Plug-In Data Centers
At Nvidia’s GTC in Berlin, Huang unveiled their new plug-in data center—the DSX Pod—as the key to unlocking Europe’s AI infrastructure lag. Oh, and autonomous humanoid robotics.
Nvidia CEO Jensen Huang unveiled his vision for Europe's AI economy during his keynote at Nvidia's GTC conference in Berlin on April 11, 2023. In his 90-minute address, he emphasized the importance of not just consuming AI, but also producing it, stating, "To be part of this AI economy, you don’t want to just be consuming the AI. You want to be producing the AI."
Huang introduced a modular "AI factory" called the DSX Pod, a plug-and-play solution for building AI data centers. The containerized system, equipped with Nvidia's latest compute racks, can be set up within 14 days, a significant improvement over traditional construction timelines. This rapid deployment could prove invaluable in Europe, where large-scale infrastructure projects are often slow and cumbersome.
The DSX Pod also promises to reduce idle compute time, as it becomes operational faster, according to Huang. The keynote also touched on the growing role of humanoid robotics in Europe, with Huang predicting the industry as "the next great industry." He highlighted Europe's manufacturing background as a strong foundation for this emerging sector. Nvidia is also partnering with European carmakers Volkswagen and Mercedes to integrate AI into autonomous vehicles.
Huang stressed that the future will see a mix of closed and open models for AI usage, with customized open models becoming increasingly prevalent. He showcased Nvidia's open-source models, which perform at frontier levels when fine-tuned with an organization's proprietary data. The CEO also predicted that agents, capable of reasoning, iterating, and using tools, will lead to higher demand for computing resources, further driving the need for powerful GPUs like those offered by Nvidia.
Written by urgent.news from Wired Business's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.