Denvr AI Cloud

Denvr AI Cloud offers a range of features including self-service provisioning of AI compute clusters, regional data centers for data residency, end-to-end encryption, and SOC 2-compliant controls. It provides high-performance GPU clusters with NVIDIA H200, H100, A100, and Intel Gaudi2 accelerators, scalable from single nodes to 1,024 GPU clusters connected by 3,200 Gbps InfiniBand. Integrated networking and storage include non-blocking InfiniBand training fabrics, tenant-isolated VPC networking, and petabyte-scale shared filesystems. The platform is built to scale with modular data center architecture, from 1 MW sites to 100+ MW campuses.

Denvr AI Cloud is used for AI training and inferencing workloads. It supports deploying foundation or custom models on dedicated GPU endpoints that are fully managed and OpenAI API compatible. The platform includes a developer platform with packaged AI applications like Jupyter Notebooks and OpenWebUI, ML-ready machine images with current GPU drivers, Docker, and popular AI frameworks. It offers REST APIs, CLI tools, and language bindings for Python and Terraform. Denvr also has a startup program providing up to $500,000 in AI cloud credits for early-stage companies, and supports Canadian AI communities through sponsorships and contributions to open infrastructure standards.

Denvr AI Cloud offers on-demand or reserved GPU compute across bare metal, VMs, and containers. The startup program provides an initial $1,000 start-up credit, with eligible teams able to receive up to $500,000 in production credits based on needs and potential. Specific pricing details are not listed on the page, but users can sign up for on-demand access or contact the team for private AI platform requirements.