GPU Cloud Provider · USA
Oblivus
Oblivus is a cloud provider that offers high-performance computing solutions with a focus on affordability and scalability. Founded in 2022, Oblivus provides a range of virtual servers and GPU instances tailored for AI, ML, and high-end computing tasks. The service features a robust global network infrastructure and is available from multiple data center regions.
GPUs
2
Founded
March 2022
Countries
4
Data Centers
4
Team Size
11-50
GPU Marketplace
NVIDIA A100 80GB SXMOn-Demand
NVIDIA H100 SXMOn-Demand
Company Profile
Company TypeStartup
Provider TypeCloud Provider
FoundedMarch 2022
HeadquartersUSA
FundingBootstrapped
Team Size11-50
Infrastructure
GPU FleetNVIDIA H100 SXM, NVIDIA H100 PCIe, NVIDIA A100 80GB, NVIDIA A100 40GB, NVIDIA RTX A6000, NVIDIA RTX 4090, NVIDIA RTX 3090
Network FabricInfiniBand, Ethernet
Connectivity400Gbps dark fiber, 200Gbps public Internet
StorageNVMe
Data Center TierCarrier-neutral colocation
Bare MetalYes
AvailabilityGA since December 2023
StartupEnterpriseResearchAI/ML Developers
Compute & Deployment
On-DemandYes
Spot / InterruptibleNo
VM-BasedYes
Container-BasedYes (Docker)
Serverless GPUNo
Spin-Up TimeUnder 2 minutes
GPU Hardware
Latest GenH100 SXM, H100 PCIe, L40S
Legacy SupportA100, A10G, RTX 4090
Multi-GPU NodesYes (up to 8x per node)
Max GPUs/Node8
NVLinkYes (SXM nodes)
InfiniBandYes (HDR 200Gbps)
PCIe vs SXMBoth PCIe and SXM
HGX PlatformYes (HGX H100 8-GPU)
Pricing Model
Per HourYes (primary billing unit)
Per MinuteNo
Spot DiscountNo spot pricing
Public PricingYes
Hidden FeesNone disclosed
Pay-as-you-goYes
Performance & Scaling
Multi-Node TrainingYes (multi-node distributed training supported with NCCL)
Elastic ScalingManual only
Auto ScalingNo
NVSwitchYes (on SXM nodes)
Perf IsolationYes (dedicated bare metal)
Noisy NeighborYes (bare metal, no sharing)
Developer Experience
OnboardingDeploy in minutes via web UI with API key-based access
FrameworksPopular machine learning and AI libraries and tools included in OblivusAI
CLI ToolingBasic CLI for instance management
JupyterVia SSH port forwarding
TemplatesPyTorch, TensorFlow, CUDA
DocumentationBasic getting-started guide with API reference
API FeaturesCLI, SDK, REST API
Security & Compliance
SecurityISO 27001,Type 2 SOC 1,Type 2 SOC 2,HIPAA regulations
ComplianceISO 27001, Type 2 SOC 1, Type 2 SOC 2, HIPAA
Data Center Locations
Coverage
CountriesUnited States, Germany, Netherlands, Finland
CitiesNew York, Frankfurt, Amsterdam, Helsinki
EuropeNorth America
Compliance Regions
EU Data ResidencyYes (Frankfurt, Amsterdam, Helsinki)
US Gov CloudNo
India RegionNo
Datacenter Locations
Key Strengths
Competitive GPU pricing for startups
Simple self-service web UI
Crypto payment support
On-demand access without waitlists
Broad GPU model selection
Known Limitations
Limited public documentation
No published SLA
Small fleet size compared to major providers
Limited support resources
No enterprise support tier
Limited geographic regions
Additional Information
Support Options
["Online Support","Dedicated 24/7 Support"]
Core Proposition
Affordable on-demand GPU cloud with no minimum commitment, targeting AI/ML workloads with competitive per-hour pricing.
Payment Methods
Credit CardCrypto
Last updated March 2026. Information subject to change.