GEN AI cluster – request access

With our state-of-the-art GEN AI cluster, we provide you with a high-performance infrastructure for demanding AI applications. The system has been specifically optimized for training, fine-tuning, and inference of the latest AI models and offers top-tier performance and maximum scalability.

The cluster comprises several high-performance nodes equipped with state-of-the-art NVIDIA H200 NVL GPUs featuring NVLink technology, as well as mid-performance nodes based on the latest RTX PRO 6000 Blackwell architecture. The environment is complemented by a high-performance InfiniBand network with speeds of up to 200 Gbit/s and a central file server with over 150 TB of storage.

This makes the cluster ideal for:

– Training small to medium-sized models–
Fine-tuning large language models with up to 70 billion
parameters– Inference on current models such as Llama 3, Qwen 3, or Gemma 3–
Vision applications such as Segment Anything or CLIP

If you would like to gain access to this infrastructure, please fill out the form below. After reviewing your request, we will contact you shortly to discuss your specific usage needs.