Cutting-edge deep learning research - DFKI is part of the NVIDIA AI Lab program

Deep Learning is the subfield of Artificial Intelligence that is currently causing considerable breakthroughs, especially in pattern recognition. This requires powerful processors that enable the training of complex neural networks with large amounts of data. Now NVIDIA, a pioneer in the field of high-performance acceleration processors, and the German Research Center for Artificial Intelligence (DFKI) are expanding their collaboration in cutting-edge research into innovative deep learning methods.

CEO Jen-Hsun Huang, NVIDIA: "The world's best AI research labs must have access to the world's most powerful AI research tools".

"The world's best AI research labs must have access to the world's most powerful AI research tools" said NVIDIA CEO Jen-Hsun Huang in his keynote address at the GPU Technology Conference in Amsterdam, announcing the expansion of the partnership with DFKI under the NVIDIA AI Lab program (NVAIL). NVIDIA already has similarly close collaborations with the Universities of Stanford and Berkeley and the OpenAI initiative. The DFKI and the Swiss AI Lab IDSIA are now the first research institutions in Europe with which NVIDIA has entered into such a partnership. The initial four-year funding for DFKI includes the provision of a DGX-1 supercomputer.

The expansion of the cooperation results from the existing close collaboration with the DFKI Deep Learning Competence Center. DFKI was last selected as a GPU Research Center by NVIDIA in June 2016 due to its outstanding research results in the field of text, image, video and social media analysis. With THE NVAIL initiative, NVIDIA aims to support the groundbreaking research of the world's leading AI scientists and institutions and to deepen the exchange with the top of the deep learning research movement.

Jaap Zuiderveld, NVIDIA vice president of Sales & Marketing EMEA, said, "NVIDIA is proud and excited to be working with DFKI to advance AI research and support the research with grants and our latest high-performance computing platform, DGX-1. We also want to provide employees and students with opportunities to help shape the future of AI computing platforms at NVIDIA".

Prof. Dr. Dr. h.c. mult. Wolfgang Wahlster, Scientific Director and CEO of DFKI: "The more hidden layers a deep learning network has, the better it can abstract and thus recognize more complex relationships and subtle features in the data. However, each layer also dramatically increases the required computing power. That's why working with NVIDIA helps us develop AI solutions for the next generation of autonomous systems, human-robot teams, and real-time controllers for Industry 4.0".  

Dr. Damian Borth, scientist in the research field of knowledge management and head of the Deep Learning Competence Center at DFKI: "Our ambition is to shape cutting-edge research in the field of machine learning while taking into account the social relevance and responsibility of AI. As the inventor of the GPU and pioneer of the Deep Learning revolution, NVIDIA is the perfect partner for this".

Prof. Dr. Andreas Dengel, head of the DFKI Research Department Knowledge Management: "The cooperation between NVIDIA and DFKI fits like a glove. It gives us access to the best AI computers. These will give us a further boost to process the large multimedia data volumes in our projects even more efficiently and effectively with the innovative network architectures we have developed".

German Research Center for Artificial Intelligence
Deutsches Forschungszentrum für Künstliche Intelligenz