Oracle Cloud

Oracle Cloud Infrastructure (OCI) provides advanced AI infrastructure designed for a wide range of AI workloads, including model training and inference. OCI supports scale-out capabilities up to zettascale with bare metal instances, equipped with the latest NVIDIA GPUs and high-bandwidth, low-latency RDMA networking.

Oracle Cloud

Provider Profile

Founded

Oracle Corporation founded in 1977, with Oracle Cloud developed later.

Headquarters

Redwood Shores, California, USA

Pricing Model

Pricing offers up to 220% better rates on GPU VMs compared to competitors, no extra charges for RDMA capability, block storage, or network bandwidth. The first 10 TB of egress is free.

Technical Specification

Target Audience
  • Enterprise ML
  • Researchers
  • Large-scale AI training
GPU Clusters & Offerings
  • NVIDIA GB200 NVL72
  • NVIDIA Blackwell GPUs
  • NVIDIA Hopper GPUs
  • NVIDIA H100, A100, H200, B200 GPUs
  • AMD MI300X GPUs
  • Ultrafast RDMA cluster networking
  • Virtual Machine and Bare Metal GPU instances
Network Fabric
RDMA over Converged Ethernet (RoCE v2)
Connectivity Bandwidth
  • Up to 3
  • 200 Gbps of cluster network bandwidth
  • up to 400 Gbps front-end network bandwidth
Storage Architecture
  • NVMe SSD
  • Oracle-managed file storage with Lustre
  • Block storage
  • Object storage
Compute Framework Compatibility
OCI Data Science
Resource Orchestration
Support for various cluster management and orchestration technologies inherent to cloud services.
Security Infrastructure
Complies with industry-standard security and compliance protocols suitable for government and dedicated clouds.
Developer Interface & APIs
  • CLI
  • SDK
  • REST API
Support Operations
  • Dedicated support
  • online resources
Resource Availability
GA
Datacenter Locations
Regulatory Compliance
Data residency, security, and compliance capabilities are supported, particularly for government and sovereign cloud services.
Key Platform Features
  • Up to 131,072 GPUs for zettascale performance
  • Cluster networks with 8X scalability
  • Unique GPU bare metal instances
  • High-performance RDMA cluster networking with latency as low as 2.5 microseconds
  • High storage limits with options for NVMe SSD and Lustre file systems

Last Audit: February 2026