How DeepL and NVIDIA are advancing Language AI infrastructure in Europe

Key Takeaways

• DeepL and NVIDIA share a strategic infrastructure partnership built on coordinating hardware, software, and optimization to advance Language AI performance.
• DeepL became the first company in Europe to deploy the NVIDIA DGX SuperPOD with DGX GB200 systems: a landmark in European AI infrastructure.
• The SuperPOD handles trillion-parameter models, enabling DeepL researchers to test ideas faster, train models more quickly, and deliver near real-time inference.
• DeepL's collaboration with NVIDIA extends beyond hardware to include joint work on LLM capability planning and supercomputer optimization algorithms.
• Optimization work at the software level unlocks meaningful gains in energy efficiency, letting DeepL do more with its expanded compute capacity.
• DeepL partnered with EcoDataCenter to install liquid cooling infrastructure in advance, making it ready to deploy NVIDIA's most powerful machine to date.
• DeepL's Language AI suite is built on infrastructure designed for enterprise-scale speed, accuracy, and multilingual performance.

AI technology requires an incredible synergy of software and computing power, unlike previous generations of SaaS products. It requires businesses that can carefully coordinate efforts, anticipate what’s necessary to advance AI, and invest in the future together.

This is exactly the type of relationship that DeepL has built with NVIDIA. It’s a relationship we celebrated at NVIDIA GTC 2025, the GPU technology conference that’s commonly considered the Woodstock of AI. 

Several members of DeepL’s technology leadership team attended GTC, including our Chief Technology Officer (CTO) Sebastian Enderlein and VP Research Stefan Mesken. 

DeepL’s Research High-Performance Computing (HPC) Engineer Markus Schnös presented the story of how our two businesses worked together to deliver the industry-leading training and inference performance of our next-gen LLMs. 

It’s a story about what it takes to push the boundaries of AI performance today. That means not only investing in the right hardware but also finding innovative ways to use it to its full potential.

Visit DeepL AI Labs to see how we’re forging the future of Language AI.

Why DeepL’s next AI infrastructure upgrade matters

It's also a story that’s just getting started. DeepL has become the first company in Europe to deploy the NVIDIA DGX SuperPOD with DGX GB200 systems. 

NVIDIA’s Chief Solutions Architect Thomas Schoenemeyer describes this cluster of NVIDIA GB200 Grace Blackwell Superchips as the AI equivalent of a championship-winning Formula 1 car. 

In NVIDIA’s previous testing of the SuperPOD’s capabilities, it led on every benchmark in the open-source MLPerf framework. That framework measures the performance of hardware for training and inference. 

The SuperPOD is capable of handling trillion-parameter models, which lets DeepL researchers realize new abilities:

• Test new ideas faster
• Train models more quickly
• Deliver near real-time inference

Having an AI Formula 1 car in the garage allows us to deliver Language AI that’s more capable, more accurate, and more adaptable, every year.

It enables us to be bold and ambitious in the tasks we set for DeepL. We can take on new use cases through solutions like DeepL Voice. 

And just as before, these performance gains don’t come from just investing in the right machines. They come from building a relationship around how we use and develop those machines. 

How DeepL and NVIDIA work together to improve AI performance

Our collaboration with NVIDIA involves more than access to advanced hardware. It also includes close work on the capabilities LLMs will require and the systems that can deliver them:

• Our collaboration with NVIDIA involves discussions about the types of training and inference capabilities that LLMs will demand.
• It includes conversations about the configuration of machines, such as the new DGX SuperPOD, that can deliver them.
• It also means working together on optimization algorithms for the software running on supercomputers.
• That work can unlock major gains in energy efficiency and let us do more with the extra capacity we have.

Anyone who watches Markus’s presentation gets great insights into what that means in practice.

Meet the new DeepL Translator: a smarter way of working with language.

Why infrastructure readiness matters for AI innovation

Finally, and crucially, a close working relationship with NVIDIA helps inform our collaboration with other suppliers. 

EcoDataCenter houses DeepL’s existing supercomputers and is one of the most advanced and sustainable data centers in the world. It had anticipated that the next generation of machines would require liquid cooling to deliver their step-change in performance.

We worked with EcoDataCenter to put the infrastructure required for this liquid cooling in place, ready to deploy the most powerful machine NVIDIA has yet built. 

That’s why DeepL was ready to become the first company in Europe to leverage the new SuperPOD. This ability to anticipate the elements of the AI ecosystem’s needs helps us keep pushing the boundaries of what this technology can do.

We’re honored to have presented at GTC in 2025. We look forward to the exciting innovations that will come from our ongoing collaboration with NVIDIA.

Read the independent report benchmarking DeepL Voice against competitors in real-time AI-translated captions.

Bring this infrastructure to your business 

This kind of investment—in partnerships, hardware, and optimization—is what powers DeepL's Language AI suite. We build translation, writing, and communication tools for speed, accuracy, and enterprise-scale performance.

Whether your teams are localizing content, improving business writing, or enabling global meetings, that infrastructure is what lets you communicate clearly and confidently across languages.

Contact Sales to see what Translator, Write, Voice, API, and Integrations can do for your organization's multilingual workflows.