OVHcloud GPU

OVHcloudCompute

Dedicated GPU instances with NVIDIA H100 / H200 / L40s / L4 card options for AI/ML training and inference workloads, hosted in OVHcloud's European data centres for sovereign-AI use cases

FluffyStack tools
Add to Service BuilderAdd to CompareCompare with equivalentsExplore OVH in TreemapExplore compute in HoneycombSee OVH regions on the World MapSee compute as a networkScore jurisdiction exposure
DocumentationPricingOVH website

Jurisdictional exposure

Provider HQ
EURoubaix, France

Subject to GDPR, Data Act, DGA

Region locations
APACEUUKUSOther14 regions across 5 jurisdictions
Sovereign option
Yes — 8 sovereign-flagged regions available
Full scorecard for this service →EU lens detail →Sovereign cloud coverage map →

Attributes

GPU Support
Yes

Sub-services (2)

NVIDIA H100

H100 SXM5 GPU instances for foundation-model training

NVIDIA H200

H200 GPU instances for the largest training and inference workloads

Compliance & Certifications

This service is attested for the following frameworks. Always verify with the provider before relying on a specific compliance posture.

GDPRISO 27001G-Cloud

Where this runs

14 regions
9 countries
8sovereign
Sovereign regions (8)
  • GravelinesOVHcloud Sovereign Cloud (EU)
  • RoubaixOVHcloud Sovereign Cloud (EU)
  • StrasbourgOVHcloud Sovereign Cloud (EU)
  • ParisOVHcloud Sovereign Cloud (EU)
  • FrankfurtOVHcloud Sovereign Cloud (EU)
  • LondonOVHcloud Sovereign Cloud (UK)
  • WarsawOVHcloud Sovereign Cloud (EU)
  • Erith · LondonOVHcloud Sovereign Cloud (UK)
Commercial regions (6)

North America (3)

  • Beauharnois
  • Hillsboro
  • Vint Hill

Asia (2)

  • Mumbai
  • Singapore

Oceania (1)

  • Sydney

Tags

gpuh100h200l40sl4ai-accelerationcomputeeuropeGDPR

Equivalent services on other platforms

Aruba GPU on DemandAruba

Hourly-billed GPU-accelerated instances with NVIDIA card options for AI/ML training and inference workloads — Italian data residency for sovereign AI workloads

Amazon EC2AWS

Secure, resizable virtual servers in the cloud with on-demand, reserved, spot, and savings-plan pricing, over 600 instance types, and Nitro-based hardware isolation

Azure Virtual MachinesAzure

On-demand scalable virtual machines with D, E, F, L, M, and N series sizes, spot pricing, reserved instances, availability sets, and availability zones

Compute EngineGCP

Scalable high-performance virtual machines with custom machine types, live migration, per-second billing, and sustained-use and committed-use discounts

IONOS Cloud CubesIONOS

Pre-configured dedicated compute instances with fixed CPU, RAM, and SSD allocations billed on a flat monthly rate, positioned as the predictable-cost alternative to the per-minute Compute Engine pricing model

Nscale GPU ClustersNscale

Dedicated NVIDIA H100 / H200 GPU clusters with InfiniBand interconnect for distributed AI training and large-model inference, positioned as the European hydropower-cooled alternative to US-hyperscaler GPU capacity

OCI ComputeOracle

Bare-metal and virtual machine compute instances on Oracle Cloud Infrastructure — Intel/AMD x86 plus Ampere A1 ARM shapes, GPU instances (NVIDIA A10/A100/H100/RTX PRO), per-second billing, and an always-free tier covering 4 OCPU ARM allocation

Outscale Flexible GPUsOutscale

GPU-accelerated compute attached to Flexible Compute Unit (FCU) VMs, with NVIDIA Tesla / Ampere card options for AI/ML training and inference workloads on SecNumCloud-eligible infrastructure

Scaleway InstancesScaleway

Cloud virtual machines including ARM-based and dedicated CPU options, billed per minute

Pricing

Pricing model:per-gpu-hour