SoftwareLore

Software profile Cloud & Infrastructure

Oracle Cloud Infrastructure

Oracle’s public cloud for computing, databases and large AI clusters

Oracle Cloud Infrastructure (OCI) is Oracle’s public cloud platform. Launched in 2016, it rents out computing, storage, networking, managed databases and the large GPU clusters used to train and run AI models.

Made by 02 1977 Oracle Data & AI Oracle Founded 1977 · Austin, Texas, United States Company profile

Oracle Cloud Infrastructure at a glance

Developer
Oracle
First released
2016Became generally available on October 20, 2016, as Oracle Bare Metal Cloud Services
Pricing model
Usage-based
Platforms
Cloud, Web
Deployment
Public cloud, Dedicated Region in customer data centers, Sovereign and government regions, Multicloud
License
Proprietary
Official site
oracle.com/cloud

What is Oracle Cloud Infrastructure?

Oracle Cloud Infrastructure, usually shortened to OCI, provides the same basic building blocks as other large public clouds: virtual machines and bare metal servers, block and object storage, virtual networks, Kubernetes, serverless functions and a catalog of more than 200 services. Oracle says it operates more than 50 public cloud regions in 28 countries, and customers pay for what they use or commit to spending in advance through Universal Credits, which Oracle prices the same in every region.

Oracle designed OCI with a new engineering team that included Clay Magouyrk, hired from Amazon Web Services in 2014, and began promoting it in 2018 as its Generation 2 cloud. The design moves network virtualization off customers’ servers, and its cluster networking uses RDMA over Converged Ethernet so that large numbers of GPUs can exchange data with low latency, a capability that became central once AI developers began renting GPU clusters at scale.

AI workloads have transformed OCI’s scale. OpenAI’s flagship Stargate campus in Abilene, Texas, runs on OCI, and in July 2025 the companies agreed to develop up to 4.5 gigawatts of additional capacity. Oracle’s cloud infrastructure revenue rose 77 percent to $18.1 billion in the fiscal year ended May 2026 and more than doubled year over year in the following quarter. OCI also hosts Oracle’s multicloud database services inside Microsoft Azure, Google Cloud and AWS, and Oracle installs complete OCI regions in customers’ own data centers through Dedicated Region.

Key features of Oracle Cloud Infrastructure

  1. 01

    Compute and bare metal servers

    Virtual machines and dedicated bare metal servers run on x86, Arm and GPU hardware, and flexible shapes let customers choose the exact number of processor cores and amount of memory they pay for.

  2. 02

    AI superclusters

    OCI Supercluster and Zettascale10 connect large numbers of NVIDIA GPUs with low-latency RDMA networking. Oracle says Zettascale10 clusters can reach up to 800,000 GPUs spread across multiple data centers.

  3. 03

    Managed database services

    Oracle AI Database runs on Exadata-based services or as Autonomous AI Database, while MySQL HeatWave, PostgreSQL and NoSQL services cover other database needs.

  4. 04

    Multicloud database services

    Oracle Database@Azure, Oracle Database@Google Cloud and Oracle Database@AWS place OCI-managed Exadata systems in partner data centers, so Oracle databases can sit next to applications running on another cloud.

  5. 05

    Distributed and sovereign cloud

    Dedicated Region installs a full OCI region inside a customer’s data center, Oracle Alloy lets partners offer OCI under their own brand, and sovereign and government regions address data residency requirements.

  6. 06

    Generative AI services

    OCI Generative AI offers hosted large language models from providers such as Cohere, Google and xAI through an API, along with tools for building retrieval-augmented applications and AI agents.

  7. 07

    Always Free tier

    Every account includes Always Free resources, such as small AMD and Arm-based virtual machines, block and object storage and Autonomous AI Database instances, plus trial credits during the first 30 days.

Who uses Oracle Cloud Infrastructure?

  • AI companies rent large GPU clusters to train and serve foundation models, often under multiyear capacity contracts.
  • Enterprises move Oracle Database, E-Business Suite and other Oracle workloads to OCI to run them on Exadata-based managed services.
  • Governments and regulated industries use sovereign regions or Dedicated Region to keep sensitive data inside national borders or their own facilities.
  • Companies standardized on Azure, Google Cloud or AWS run Oracle databases inside those clouds through Oracle’s multicloud database services.
  • Consumer internet services such as Zoom and TikTok’s U.S. operations use OCI for large-scale computing and data hosting.

History of Oracle Cloud Infrastructure

Oracle began assembling a new cloud engineering team in 2014, recruiting engineers from established cloud providers, including Clay Magouyrk from Amazon Web Services. The service became generally available on October 20, 2016, as Oracle Bare Metal Cloud Services, and in 2018 Oracle renamed it Oracle Cloud Infrastructure and began promoting it as a Generation 2 cloud.

OCI gradually won large workloads beyond Oracle’s traditional database customers, including Zoom, which chose it in 2020 during the surge in video meetings. Oracle added an Always Free tier in 2019, full cloud regions for customer data centers in 2020 and, after an interconnect partnership with Microsoft, Oracle Database@Azure in 2023. Generative AI then made GPU capacity OCI’s main growth engine: Oracle joined OpenAI and SoftBank in the Stargate project announced in January 2025, and by fiscal 2026 its cloud infrastructure revenue had overtaken its cloud applications revenue.

  1. 2016

    Oracle Bare Metal Cloud Services becomes generally available on October 20.

  2. 2018

    Oracle renames the service Oracle Cloud Infrastructure and markets it as a Generation 2 cloud.

  3. 2019

    Oracle launches its Cloud Free Tier with Always Free services and Autonomous Database instances.

  4. 2020

    Zoom selects OCI to help handle rapid growth in video meetings.

  5. 2023

    Oracle Database@Azure brings Oracle database services into Microsoft’s data centers.

  6. 2025

    OpenAI and Oracle agree to develop up to 4.5 gigawatts of additional Stargate data center capacity.

  7. 2026

    OCI revenue reaches $18.1 billion for fiscal 2026, up 77 percent from the year before.

Oracle Cloud Infrastructure pricing

Pricing modelUsage-based

Pay-as-you-go or prepaid Universal Credits, billed by usage at the same prices in every region, with an Always Free tier and 30 days of trial credits for new accounts.

Oracle Cloud Infrastructure alternatives

Well-known alternativesAmazon Web Services, Microsoft Azure, Google Cloud, CoreWeave and IBM Cloud

Oracle Cloud Infrastructure: frequently asked questions

Who owns Oracle Cloud Infrastructure?

Oracle Cloud Infrastructure is owned and operated by Oracle Corporation. Clay Magouyrk, who helped build OCI after joining Oracle from Amazon Web Services in 2014 and later became its president, was named one of Oracle’s two chief executives in September 2025, alongside Mike Sicilia.

What is Oracle Cloud Infrastructure used for?

Oracle Cloud Infrastructure is used to run applications, databases and AI workloads in Oracle’s data centers instead of on company-owned servers. Common uses include training and serving AI models on large GPU clusters, hosting Oracle Database and Exadata workloads, running enterprise applications and meeting data sovereignty rules with dedicated or government regions.

Is Oracle Cloud Infrastructure free?

OCI has a Free Tier that includes Always Free services, such as small virtual machines, storage and Autonomous AI Database instances, with no time limit, plus trial credits for the first 30 days. Beyond those limits, customers pay for the resources they use on a pay-as-you-go basis or through prepaid Universal Credits.

When was Oracle Cloud Infrastructure released?

Oracle’s current cloud platform became generally available on October 20, 2016, under the name Oracle Bare Metal Cloud Services. Oracle renamed it Oracle Cloud Infrastructure in 2018 and has since expanded it to more than 50 public cloud regions in 28 countries.

How is Oracle Cloud Infrastructure connected to OpenAI and Stargate?

OpenAI is one of OCI’s largest customers. The flagship Stargate AI data center in Abilene, Texas, runs on OCI, and in July 2025 OpenAI and Oracle agreed to develop up to 4.5 gigawatts of additional capacity in the United States, a partnership OpenAI said was worth more than $300 billion over five years.

Sources

  1. Oracle announces Q1 fiscal 2027 results oracle.com
  2. OpenAI, Oracle and SoftBank expand Stargate with five new AI data center sites, SoftBank Group group.softbank
  3. Public cloud regions and data centers, Oracle oracle.com
  4. Oracle Cloud pricing oracle.com
  5. Oracle Cloud, Wikipedia en.wikipedia.org

Last reviewed . Spotted an error? Send a correction.

to move · Enter to open · Esc to close