Oracle Unveils Next-Generation Oracle Cloud Infrastructure Zettascale10 Cluster for AI

Rhea-AI Summary

Oracle (ORCL) announced OCI Zettascale10, which Oracle describes as the largest AI supercomputer in the cloud, delivering up to 16 zettaFLOPS of peak performance by connecting hundreds of thousands of NVIDIA GPUs across multi–gigawatt clusters.

OCI Zettascale10 uses Oracle Acceleron RoCE networking and NVIDIA AI infrastructure, targets initial deployments of up to 800,000 NVIDIA GPUs, is being deployed with OpenAI at the Stargate site in Abilene, Texas, and will be available in the second half of next calendar year. Oracle says the design prioritizes ultra‑low GPU‑to‑GPU latency, cluster utilization, reliability, and power efficiency.

Positive

• 16 zettaFLOPS peak performance
• Targets up to 800,000 NVIDIA GPUs per deployment
• Joint supercluster with OpenAI at Stargate in Abilene, Texas
• Designed for ultra‑low GPU‑to‑GPU latency and higher cluster utilization

Negative

• Operates in multi‑gigawatt data center campuses, implying very high power demand
• Clusters concentrated within a two‑kilometer radius, increasing site concentration risk

Insights

Cloud infrastructure strategist

Oracle announces a large-scale OCI Zettascale10 supercluster with up to 16 zettaFLOPS and deployments up to 800,000 GPUs.

OCI Zettascale10 packages massive GPU count, custom RoCE networking, and multi–gigawatt data‑center campuses to offer extremely high aggregate AI throughput. The collaboration with OpenAI at the Stargate site and the stated support for up to 800,000 NVIDIA GPUs and up to 16 zettaFLOPS signal an intent to serve hyperscale model training at cloud scale.

Key dependencies include physical campus density, network latency within the stated two‑kilometer radius, and successful delivery of the Oracle Acceleron RoCE fabric across planes without introducing new operational failure modes. Watch order intake and the cited availability window ("second half of next calendar year") for concrete deployments and revenue recognition; near‑term indicators include customer orders and initial performance validation at the Stargate site within months of availability.

AI compute architect

Oracle presents a networking‑centric approach to GPU scaling that emphasizes low latency, resilience, and power efficiency for large model workloads.

The core mechanism replaces a deeper network tiering model by using GPU NIC switching and multiple isolated planes to increase scale and avoid single‑plane congestion. This should improve predictable GPU‑to‑GPU latency and lower non‑compute power draw via Linear Pluggable Optics and Linear Receiver Optics while enabling plane‑level maintenance without full cluster interruption.

Risks include the complexity of managing many isolated network planes, verifying fabric‑wide latency at the claimed scale, and integrating customer software stacks with the custom RoCE design. Monitor technical benchmarks from the Stargate collaboration, multi‑customer deployment confirmations, and reported GPU‑to‑GPU latency numbers after early customer rollouts; expect these concrete signals in the months following initial availability.

Largest AI supercomputer in the cloud delivers 10X the amount of zettaFLOPS of peak performance

Built on Oracle Acceleron RoCE networking architecture with NVIDIA AI infrastructure , OCI Zettascale10 will provide multi–gigawatt AI workload capacity and scale

LAS VEGAS, Oct. 14, 2025 /PRNewswire/ -- Oracle AI World – Oracle today announced Oracle Cloud Infrastructure (OCI) Zettascale10, the largest AI supercomputer in the cloud. OCI Zettascale10 connects hundreds of thousands of NVIDIA GPUs across multiple data centers to form multi-gigawatt clusters that deliver up to an unprecedented 16 zettaFLOPS of peak performance. OCI Zettascale10 is the fabric underpinning the flagship supercluster built in collaboration with OpenAI in Abilene, Texas, as part of Stargate. Built on next-generation Oracle Acceleron RoCE networking architecture, OCI Zettascale10 is powered by NVIDIA AI infrastructure that delivers breakthrough scale, extremely low GPU-GPU latency across the cluster, industry-leading price-performance, improved cluster utilization, and the reliability required for large scale AI workloads.

OCI Zettascale10 is a powerful evolution of the first Zettascale cloud computing cluster, which was introduced in September 2024. OCI Zettascale10 clusters are housed in large gigawatt data center campuses that are hyper-optimized for density within a two-kilometer radius to offer the best GPU-GPU latency for large scale AI training workloads. This architecture is being deployed with OpenAI at the Stargate site in Abilene.

"With OCI Zettascale10, we're fusing OCI's groundbreaking Oracle Acceleron RoCE network architecture with next-generation NVIDIA AI infrastructure to deliver multi–gigawatt AI capacity at unmatched scale," said Mahesh Thiagarajan, executive vice president, Oracle Cloud Infrastructure. "Customers can build, train, and deploy their largest AI models into production using less power per unit of performance and achieving high reliability. In addition, customers will have the freedom to operate across Oracle's distributed cloud with strong data and AI sovereignty controls."

"OCI Zettascale10 network and cluster fabric was developed and deployed first at the flagship Stargate site in Abilene, Texas – our joint supercluster with Oracle," said Peter Hoeschele, vice president, Infrastructure and Industrial Compute, OpenAI. "The highly scalable custom RoCE design maximizes fabric-wide performance at gigawatt scale while keeping most of the power focused on compute. We're excited to keep scaling Abilene and the broader Stargate program together."

OCI plans to offer multi-gigawatt deployments of OCI Zettascale10 to customers. Initially, OCI Zettascale10 clusters will target deployments of up to 800,000 NVIDIA GPUs delivering predictable performance and strong cost efficiency, with high GPU–to–GPU bandwidth enabled by Oracle Acceleron's ultra–low–latency RoCEv2 networking.

"Oracle and NVIDIA are bringing together OCI's distributed cloud and our full–stack AI infrastructure to deliver AI at extraordinary scale," said Ian Buck, vice president of Hyperscale, NVIDIA. "Featuring NVIDIA full-stack AI infrastructure, OCI Zettascale10 provides the compute fabric needed to advance state–of–the–art AI research and help organizations everywhere move from experimentation to industrialized AI."

Oracle Acceleron RoCE networking delivers scale, reliability, and efficiency for AI on OCI Zettascale10

Oracle Acceleron RoCE networking architecture is a critical innovation for customers to build, train, and inference AI workloads in the cloud, while taking full advantage of OCI Zettascale10's power and capabilities. It uses the switching capability built into modern GPU NICs (network interface cards), allowing them to connect to multiple switches simultaneously, with each on a separate and isolated network plane. This approach dramatically increases the network's overall scale and reliability by shifting traffic to other network planes when one has a problem, avoiding costly stalls and restarts. Key features of Oracle Acceleron RoCE networking that help customers with their critical AI workloads, include:

• Wide, shallow, resilient fabric: Helps customers deploy larger AI clusters faster at lower total cost by using the GPU NIC as a mini–switch and connecting to multiple physically and logically isolated planes. This boosts scale while reducing network tiers, cost, and power.
• Higher reliability: Helps customers maintain the stability of AI jobs by eliminating data sharing across planes. This shifts traffic away from unstable or congested planes, which keeps training jobs running and avoids costly checkpoint restarts.
• Consistent performance: Provides customers with more uniform GPU–to–GPU latency by removing a tier versus traditional three-tier designs, improving predictability for large–scale AI training and inference.
• Power–efficient optics: Supports customer workloads with Linear Pluggable Optics (LPO) and Linear Receiver Optics (LRO) to cut network and cooling costs without sacrificing 400G/800G throughput. This allows customers to devote more of their power budget to compute.
• Operational flexibility: Helps customers reduce downtime and speed up feature rollouts through plane–level maintenance and independent network operating system updates.

OCI is now taking orders for OCI Zettascale10, which will be available in the second half of next calendar year, with up to 800,000 NVIDIA AI infrastructure GPU platforms.

Additional Resources

• Watch the Oracle AI World keynote with Mahesh Thiagarajan
• Learn more about OCI AI infrastructure

About Oracle
Oracle offers integrated suites of applications plus secure, autonomous infrastructure in the Oracle Cloud. For more information about Oracle (NYSE: ORCL), please visit us at oracle.com.

About Oracle AI World
Oracle AI World is where customers and partners discover the latest product and technology innovations, see how AI is being applied across industries, and connect with experts and peers. Attendees will gain practical tips and insights to drive immediate impact within their organizations and explore how Oracle is helping unlock the full potential of cloud and AI. Join the event to see new capabilities in action and hear from thought leaders and industry movers. Register now at oracle.com/ai-world or follow the news and conversation at oracle.com/news and linkedin.com/company/oracle.

Future Product Disclaimer
The preceding is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. The development, release, timing, and pricing of any features or functionality described for Oracle's products may change and remains at the sole discretion of Oracle Corporation.

Forward-Looking Statements Disclaimer
Statements in this article relating to Oracle's future plans, expectations, beliefs, and intentions are "forward-looking statements" and are subject to material risks and uncertainties. Many factors could affect Oracle's current expectations and actual results, and could cause actual results to differ materially. A discussion of such factors and other risks that affect Oracle's business is contained in Oracle's Securities and Exchange Commission (SEC) filings, including Oracle's most recent reports on Form 10-K and Form 10-Q under the heading "Risk Factors." These filings are available on the SEC's website or on Oracle's website at oracle.com/investor. All information in this article is current as of October 14, 2025 and Oracle undertakes no duty to update any statement in light of new information or future events.

Trademarks
Oracle, Java, MySQL and NetSuite are registered trademarks of Oracle Corporation. NetSuite was the first cloud company—ushering in the new era of cloud computing.

 

View original content to download multimedia:https://www.prnewswire.com/news-releases/oracle-unveils-next-generation-oracle-cloud-infrastructure-zettascale10-cluster-for-ai-302583054.html

SOURCE Oracle

FAQ

What peak performance does Oracle say OCI Zettascale10 (ORCL) delivers?

Oracle states OCI Zettascale10 delivers up to 16 zettaFLOPS of peak performance.

When will OCI Zettascale10 be available for customers from Oracle (ORCL)?

OCI Zettascale10 is expected to be available in the second half of next calendar year.

How many GPUs does Oracle target per OCI Zettascale10 deployment (ORCL)?

Initial deployments target up to 800,000 NVIDIA GPUs.

Is OCI Zettascale10 connected to any partner projects like OpenAI (ORCL)?

Yes; OCI Zettascale10 is the fabric behind the flagship supercluster built with OpenAI at the Stargate Abilene site.

What networking technology does Oracle (ORCL) use in OCI Zettascale10?

OCI Zettascale10 uses Oracle Acceleron RoCE networking architecture with ultra‑low latency RoCEv2.

What are the main operational tradeoffs for OCI Zettascale10 (ORCL)?

Oracle highlights higher performance and utilization but notes deployments run in multi‑gigawatt campuses, indicating significant power and site concentration needs.