This advancement allows DeepL to translate the entire web in just 18.5 days, a significant improvement from the previous 194 days.

DeepL's CEO, Jarek Kutylowski, emphasized the company's commitment to research and innovation, positioning it as a leader in the competitive Language AI market.

DeepL has made a significant leap in its Language AI capabilities by deploying the NVIDIA DGX SuperPOD with DGX GB200 systems in Europe, marking it as the first company in the region to do so.

This new liquid-cooled cluster allows for scalability and the training of high-performance AI models, surpassing the capabilities of DeepL's previous supercomputer, DeepL Mercury.

The upgraded infrastructure is expected to deliver 30 times more text output compared to previous systems, greatly enhancing productivity for users.

Immediate benefits of this deployment include drastically reduced translation times for complex texts, such as translating the Oxford English Dictionary from 39 seconds to just two seconds.

Additionally, the system will expedite the translation of Marcel Proust's 'In Search of Lost Time' from 0.95 seconds to 0.09 seconds.

NVIDIA's Vice President, Charlie Boyle, highlighted the importance of robust AI deployments in Europe for maintaining competitive advantage and driving innovation across various industries.

The new setup will also improve translation quality, speed, and introduce more interactive features for users, fostering a personalized experience.

In the long term, DeepL aims to explore multi-modal models and expand generative capabilities within its platform to further enhance user interaction.

DeepL serves over 200,000 businesses and millions of individuals across 228 markets, focusing on human-like translation and improved writing solutions.

Looking ahead, the company plans to launch the world's first industrial AI cloud in Germany, which will enable manufacturers to create digital twins and enhance operations through the Omniverse platform.