#Paris #France #NVIDIA #DeepL #Language AI

DeepL Pioneers the Use of NVIDIA DGX SuperPOD in Europe

DeepL, recognized as a leading force in AI language technology, has made a groundbreaking move by implementing the NVIDIA DGX SuperPOD with DGX GB200 systems, becoming the first European company to do so. This pivotal deployment not only enhances the capabilities of DeepL's language AI but is also expected to revolutionize how users experience translation technology.

On June 11, 2025, in a significant announcement made in Paris, France, DeepL revealed that the integration of this advanced computing platform will cut the translation time for the entire internet down from 194 days to just 18.5 days. This remarkable achievement signifies a swift advancement in DeepL's research and product functionalities.

The Impact of NVIDIA DGX SuperPOD

The NVIDIA DGX SuperPOD is designed with cutting-edge features, including fluid-cooled rack-scale design and scalability to accommodate tens of thousands of GPUs. This enhanced computing power allows DeepL to execute high-performance AI models essential for advanced generative applications.

With this new infrastructure, DeepL aims to improve the overall quality, speed, and nuance of translations while introducing interactive features to its AI technology platform. Users can look forward to immediate improvements in translation services, including faster and more nuanced translations alongside a greater level of interactivity.

Being the third implementation of NVIDIA DGX SuperPOD for DeepL, this rollout surpasses the capabilities of its previous flagship supercomputer, DeepL Mercury. The continued investment in research and development marks DeepL's commitment to ensuring that it remains at the forefront of the AI language technology industry.

Jarek Kutylowski, CEO and Founder of DeepL, expressed pride in their relentless pursuit of innovation, stating, "By equipping our research infrastructure with the latest technology, we improve not only our existing offerings but also create exciting new products. The pace of innovation in AI is faster than ever, making the integration of these advanced developments essential for our continued growth."

Future Expectations and User Experience

DeepL's vision for the future includes a stronger emphasis on generative functions and personalization options to provide users with tailored experiences. Moving forward, multimodal models are expected to become standard in DeepL’s offerings, enhancing how the platform serves diverse user needs.

This new technology allows unprecedented speed and accuracy in translation. Noteworthy improvements include:

• - Translation of the Oxford English Dictionary: reduced from 39 seconds to just 2 seconds.
• - Translation of Marcel Proust’s "In Search of Lost Time": shortened from 0.95 seconds to a remarkable 0.09 seconds.
• - Increased efficiency: the new clusters can output 30 times more text than before.

As highlighted by Charlie Boyle, Vice President of DGX Systems at NVIDIA, Europe requires reliable AI implementations to maintain its competitive edge. He stated, "With the power and efficiency of our latest AI infrastructure, DeepL can expedite groundbreaking developments in language AI and provide transformative experiences to users across the continent and beyond."

About DeepL

DeepL has successfully aided over 200,000 enterprises and government entities, as well as millions of individual users across 228 markets, in overcoming language barriers. Established in 2017 by Dr. Jaroslaw Kutylowski, the company continues to grow, supported by top-tier investors like Benchmark, IVP, and Index Ventures. DeepL’s commitment to security and enhanced communication solutions helps businesses expand into new markets and significantly boosts productivity. To learn more, visit deepl.com.

In a rapidly evolving technological landscape, DeepL's latest advancements reaffirm its status as a leader in language AI, marking a new era for translation technology in Europe and beyond.