Oblivus

What's good about Oblivus

• Transparent per-GPU on-demand pricing across the catalog
• Multiple H100 SKU options (SXM5, NVLink, PCIe)
• RTX-class GPUs available for cost-sensitive workloads
• ISO 27001, Type 2 SOC 1, Type 2 SOC 2 and HIPAA compliance at both multi-tenant sites
• Public API for instance deployment and management

Oblivus GPU prices

7 GPU models across 7 configurations. Rank compares Oblivus's cheapest on-demand listing for each GPU with every provider's cheapest.

What a run costs on Oblivus

24 hours at today's cheapest listing for each GPU, at the size it is listed.

• 1× H100 SXM for 24 hours

  At $4.31/GPU-hr

  $103.49

  $26.69 over the market median rate

• 1× H200 for 24 hours

  At $3.99/GPU-hr

  $95.76

  $13.20 under the market median rate

• 1× RTX 4090 for 24 hours

  At $0.640/GPU-hr

  $15.36

  $5.16 over the market median rate

The market median is every provider's cheapest on-demand price for that GPU.

Billing & terms

On-Demand

Billed per minute while an instance runs, with multi-GPU pricing scaling linearly from per-GPU rates

Reservations

Reserved rates published per GPU type, roughly a quarter to a third below the on-demand rate

Regions

Nine regions with multi-tenant and single-tenant options; multi-tenant sites are MON1 in Montreal, Canada (Hopper, Ampere and Ada Lovelace GPUs) and OSL1 in Oslo, Norway (Ampere and Ada Lovelace GPUs)

Support

Documentation, FAQ and a ticket-based support system

Watch out for

• Smaller global footprint than hyperscalers
• Multi-tenant capacity is limited to two locations (Montreal and Oslo)
• Less established brand than the largest neoclouds
• Blackwell data-center GPUs are not in the public catalog; H200 is by reservation request and GH200 by waitlist

Features

Single-GPU Pricing

Per-GPU rates published for each instance type, with multi-GPU pricing scaling linearly

Multiple H100 Variants

H100 SXM5, NVLink and PCIe SKUs published separately for tradeoffs in interconnect

Reservation Discounts

Reserved rates are published beside the on-demand rate for each GPU type

OblivusAI Images

Ubuntu 22.04-based image preloaded with common machine learning and AI tools and libraries

Compliance Certifications

Multi-tenant data centers are ISO 27001, Type 2 SOC 1, Type 2 SOC 2 and HIPAA compliant

No-Cost Data Transfer

200Gbps+ public internet connectivity and 400Gbps+ dark fiber between data centers, with data transfers not charged

Getting started

1. Step 1

   Create an account

   Sign up at the Oblivus console

2. Step 2

   Pick a GPU and region

   Choose an on-demand instance or request a reservation

3. Step 3

   Launch and connect

   Provision via console or API and connect over SSH