3 Minutes
DeepL Sets a New Benchmark in AI-Powered Language Translation
Artificial intelligence continues to redefine the boundaries of technological innovation. DeepL, a leading machine translation company, has announced a groundbreaking achievement: the ability to translate the entire content of the internet into multiple languages in just 18.5 days. This dramatic leap forward shatters the previous benchmark of 194 days, marking an impressive 90% acceleration in large-scale translation speed.
Behind the Record: AI Supercomputing with NVIDIA DGX SuperPOD
DeepL's success is rooted in its strategic adoption of cutting-edge server technology. The company credits its rapid translation breakthrough to the deployment of the NVIDIA DGX SuperPOD supercomputer, featuring state-of-the-art DGX GB200 chips. Hosted at Sweden’s EcoDataCenter, this high-performance infrastructure represents DeepL's third major collaboration with NVIDIA—the renowned 'green team' of the AI hardware world. This infrastructure allows DeepL’s linguistic artificial intelligence platform to scale at speeds and levels of accuracy previously thought unattainable.
Leadership Perspective and Focus on Innovation
Jarek Kutylowski, DeepL’s founder and CEO, attributes the company’s advantage to relentless investment in research and development. "Staying ahead in the fast-evolving AI translation market requires continuous innovation. Upgrading with the latest technologies not only enhances our current services but also paves the way for revolutionary new products," Kutylowski observed. He emphasized that ongoing advances in AI make robust infrastructure upgrades vital for sustainable growth.
Product Performance: Blazing Speed and Scalability
Estimating the true volume of online data is a challenge, given diverse sources and ever-evolving metrics. Yet, DeepL is undaunted in its ambitions. The company also asserts that it can translate the entire Oxford English Dictionary in just two seconds and process Marcel Proust's famously lengthy novel "In Search of Lost Time" in less than 0.09 seconds—faster than the blink of an eye. Such performance highlights the leap forward in both processing power and AI algorithm sophistication.
Use Cases, Features, and Market Impact
While the primary spotlight falls on translation speed, DeepL’s use of the NVIDIA DGX SuperPOD also empowers its R&D teams to develop more nuanced linguistic tools. Enhanced computational muscle enables deeper understanding of textual subtleties, ultimately driving more accurate, context-aware translations. This means tailored solutions for complex, specialized business needs, giving DeepL a competitive edge in a crowded AI translation market.
European Relevance and Regulatory Compliance
Notably, just 10 days after NVIDIA CEO Jensen Huang's keynote at GTC Paris—where he emphasized the strategic need for European AI data centers—DeepL has become the first to deploy the DGX SuperPOD with DGX GB200 chips in Europe. The move comes at a crucial time: as European companies face rigorous data privacy regulations like GDPR and the EU AI Act, high-performing local infrastructure is increasingly essential. DeepL’s deployment not only accelerates AI innovation across Europe but also helps enterprises remain compliant with stringent privacy mandates.
Source: deepl
Comments