GPU Marketplace
Estimates only — rates are collected automatically from public provider pages and may be out of date. Prices vary by region, commitment term, and availability, and typically exclude storage, egress, and tax. Confirm current pricing with the provider before purchasing. Last collected .
Company Profile
Company TypeSovereign AI Cloud on NVIDIA GPUs
Provider TypeSovereign AI Cloud on NVIDIA GPUs
Legal EntityTaiga Cloud
Infrastructure
GPU FleetNVIDIA H100 SXM, NVIDIA H200 SXM, nodes listed with 8 GPUs
Total GPU Capacity24,000 GPUs
Bare MetalBare metal-as-a-service
AI pioneers
Compute & Deployment
On-DemandPower your projects with on-demand services
Bare MetalBare metal-as-a-service
Container-BasedManaged Kubernetes support
Kubernetesscaling with Kubernetes or SLURM
TerraformTerraform support
GPU Hardware
Latest GenNVIDIA H100/H200 SXM GPUs
Multi-GPU Nodes8
Max GPUs/Node8
Pool Size24000
PCIe vs SXMSXM
Pricing Model
Per HourBare metal NVIDIA H100 SXM: from $2.50 $/GPU/h; Bare metal NVIDIA H200 SXM: from $2.95 $/GPU/h; On-demand NVIDIA H100 SXM: from $2.70 $/GPU/h; On-demand NVIDIA H200 SXM: from $3.10 $/GPU/h
Per Minuteper-minute billing
Subscriptiontransparent pay-as-you-go, per-minute billing
Public Pricinghttps://northerndata.de/taiga-cloud-ai-pricing
Pay-as-you-gotransparent pay-as-you-go, per-minute billing
Performance & Scaling
Multi-Node TrainingGPU-as-a-service: Unlock instant access to GPUs tailored for rapid prototyping, training, and scaling with Kubernetes or SLURM.
Elastic Scalingscale with confidence and drive innovation without compromise.
Auto ScalingBuilt-in autoscaling for efficient workload management
Perf IsolationPredictable costs and guaranteed resource availability for your workloads.
Developer Experience
OnboardingSelf-service, automated platform offering AI compute on demand; immediate, flexible access; transparent pay-as-you-go, per-minute billing; Instant availability; Pre-configured environments
Templatespre-built templates
Documentationpre-built templates; API-first architecture; seamless model integration; Fully optimized AI stacks (NIM, NeMo, Triton, RAPIDS); Full observability; autoscaling; Terraform support
Security & Compliance
24000 GPUs6 AI islands in five low latency AI hubs970 PFlop/s AI Linpack performance (Rmax)NVIDIA H100/H200 SXM GPUscase studies with Noxtua and NUA Software AIpartnership with Ardent Data Centers (150+ kW racks)
Data Center Locations
Coverage
CountriesUK, Portugal, Sweden, Norway, Netherlands
CitiesChester, Sines, London, Boden, Kristiansand, Lefdal, Amsterdam
6 AI islands in five low latency AI hubs worldwide
Compliance Regions
Datacenter Locations
Key Strengths
Vertical integration and continuous optimization of operational costs
backed by one of Europe’s largest dedicated AI clouds
large fleet (24
000 GPUs) and global AI hubs (6 AI islands in five low latency AI hubs)
high AI Linpack performance (970 PFlop/s Rmax)
offers bare metal-as-a-service and GPU-as-a-service
managed Kubernetes and SLURM
pre-configured environments with autoscaling, optional private networking, and auto-healing nodes
per-minute pay-as-you-go billing
Known Limitations
No spot instance pricing mentioned
no SLA uptime percentage provided
no SDK language list or Jupyter support mentioned
no free tier or payment methods listed
no reserved instance details
no data center tier or PUE specified
Additional Information
Community
Insights (blog / updates)
Green Energy
Sustainable infrastructure and sustainable, high-speed compute locations
Core Proposition
Vertical integration and continuous optimization of operational costs allows Taiga Cloud to deliver predictable and competitive pricing that maximizes value for your AI workloads.
Notable Customers
Noxtua
NUA Software AI
Last updated August 2026. Information subject to change.