09:15
DeepL deploys NVIDIA DGX SuperPOD to expand its LLM capabilities
Germany-based neural machine translation service DeepL announced that it has deployed NVIDIA's AI data center infrastructure platform DGX SuperPOD in one of its data centers in Sweden to expand its LLM capabilities. According to DeepL, this is "the first commercial deployment of this magnitude in Europe". Benchmarks show a performance of 21.85 PFlop/s, which ranks 26th globally and 8th in Europe. Consisting of 68 NVIDIA DGX H100 systems, the NVIDIA DGX SuperPOD will help DeepL train large language models faster and develop new AI communication tools for global markets.

