AI Services
evroc is building AI Services for Europe — powerful, scalable, and secure. We offer a full-stack solution with GPU-optimized clusters, on-demand VMs, and AI services through our Think platform.
Register if you're interested in reserving capacity or get in touch.
AI Services
GPU clusters
We provide Nvidia Blackwell B200 GPU clusters with fully optimized 3.2 Tbit/s networking for training and inference workloads. You get bare metal access to GPU nodes with the latest essential Nvidia software components pre-installed. We offer Fast file storage optimized for AI workloads. We support you to get going quickly with latest Pytorch builds for Blackwell. We can optionally provide Slurm or Kubernetes orchestration on clusters.
€12 per hour for B200. Contact us for long term commitments.
VMs with GPUs
As an integrated part of the evroc VM compute service, we provide Nvidia L40S and B200 GPUs, in configurations from 1 to 4 (8) GPUs per VM. The GPU VMs come with the latest essential Nvidia software components pre-installed.
€12 per hour for B200. Contact us for long term commitments.
Think Services
The evroc Think services go beyond infrastructure and offer AI development and operational services. In Think Studio, a GPU-powered Python Notebook service, data scientists can train AI models and develop AI services using Nvidia H100 and B200 GPUs. In Think Models, a Model-as-a-service inference platform, we host leading open source AI models at dedicated API endpoints.
Private beta
The European Cloud
