We are thrilled to announce the immediate availability of our next-generation Oracle Exadata X11M, providing the choice and flexibility to deploy everywhere customers need–on-premises, Cloud@Customer, Oracle Cloud, and multicloud. Exadata is used by thousands of organizations, including the majority of the world’s largest financial, telecom, and retail businesses, to run their most critical and demanding Oracle Database workloads. The 13th generation of Exadata builds on decades of engineering excellence to power mission-critical AI, analytics, and OLTP workloads globally. Offered at the same price as the prior generation, Exadata X11M provides extreme performance, scalability, and availability for all Oracle Database workloads.
Exadata X11M Delivers Greater Value
Exadata X11M is the latest generation of the Exadata family of products and services. Featuring faster state-of-the-art hardware components and our next-generation intelligent data architecture, Exadata X11M is now available on-premises with Exadata Database Machine, as a hybrid cloud with Exadata Cloud@Customer, and in Oracle Cloud Infrastructure (OCI). Exadata X11M will be available through our multicloud partners, Microsoft Azure, Google Cloud, and AWS, making this exciting generation available anywhere you need it! What’s more, Exadata X11M is available at the same price as our previous generation, delivering greater value to our customers through accelerated performance, increased efficiency and sustainability, and Oracle Database 23ai so they can develop their applications anywhere, run them everywhere, and do more for the same price.
AI Vector Search on Exadata X11M with the latest Exadata System Software accelerates AI workloads by up to 30X by offloading processing to the storage servers. This saves significant amounts of data from being sent across the network and takes advantage of the CPUs located right next to the data. Exadata System Software also includes optimizations that track and adaptively filter the top-K results for queries offloaded to the storage, reducing data sent to the database servers by up to 4.7X. In addition, Exadata calculates and sends the result of the vector distance function back to the database server, avoiding additional CPU usage on the database servers and accelerating queries by up to 4.6X. Finally, AI Vector queries are up to 32X faster queries when using the BINARY Vector dimension formats in Oracle Database 23ai.
These same workloads gain significant performance benefits with Exadata X11M utilizing the faster CPUs and memory in the database servers. AI queries that utilize In-Memory Neighbor Graph Vector Index (aka Hierarchical Navigable Small World (HNSW)) indexes get up to 43% faster than X10M. AI queries using Neighbor Partition Vector Index (aka Inverted File Flat (IVF)) on storage servers are up to 55% faster!
For mission-critical workloads, such as OLTP, Exadata X11M uses the latest AMD EPYC processor cores, which enable up to 1.25X more concurrent transactions and up to 1.25X faster serial transactions. Write IOPS increases up to 9% on the storage servers to 1,000,000. Read latency from Exadata RDMA Memory (XRMEM) gets even faster, dropping from 17 microseconds to as low as 14 microseconds – an incredible 21% faster than the previous generation. Finally, single block reads from flash get up to 43% faster.
Analytics workloads also benefit from faster processors in the database and storage servers, with up to 25% faster analytic query processing on the database servers after Smart Scan and other storage server optimizations have returned results for further/final processing. Regarding Smart Scan, Exadata System Software 24ai on Exadata X10M added new capabilities for analytics queries to utilize XRMEM and flash. Smart Scan intelligently accesses data in both flash and XRMEM to satisfy analytic queries. By doing so, the SQL Scan Throughput from XRMEM of an X10M High Capacity storage server went from 45GB/s to 380GB/s. In X11M, this number leaps to 500GB/s per storage server – both High Capacity and Extreme Flash. The flash SQL Scan Throughput in X11M HC and EF storage is 100 GB/s – a 2.2X increase over the 45GB/s mentioned above. The storage servers also accelerate offloaded analytics queries thanks to faster processors with up to 11% faster cores. The combined effect of the faster cores and the faster flash and memory scan throughput significantly increases the overall analytic query throughput of the system.
Sustainability and Efficiency
The need for our customers to address sustainability is more urgent than ever. They must minimize environmental impact, reduce operational costs, and navigate a complex regulatory landscape. Keeping up with today’s sustainability demands, Exadata X11M helps customers reduce their power consumption and costs. Exadata X11M’s extreme performance enables customers to efficiently consolidate even more Oracle Database workloads on fewer systems, saving on infrastructure, power, cooling, data center, and operational costs. As IT executives are constantly faced with increasing workload demands and less budget to execute with, Exadata X11M is the obvious choice to do more with less.
Exadata X11M also includes new intelligent power efficiency capabilities to limit the power consumption of database server CPUs to a specific power target or enable the database server to automatically and dynamically save power when workload demand is low, for example, on weeknights and weekends. Exadata X11M also intelligently turns off unneeded cores further enabling customers to conserve energy.
• Run more databases and host more users • Run more workloads and store more data • Run more AI, analytics, and transactions • Or, run the same workload using less hardware
Exadata X11M lets you spend less!
Customers taking advantage of Oracle Database services on Exadata Cloud@Customer and in the public cloud in OCI or with multicloud services in Microsoft Azure, Google Cloud, and AWS benefit from greater operational efficiency due to built-in automation and tooling. Customers can achieve even greater efficiency by using Oracle Autonomous Database, which helps eliminate manual database management tasks and human error.
Exadata Exascale Intelligent Data Architecture for the Cloud Era
Exadata Exascale is the world’s only intelligent data architecture for the cloud. Exascale is a revolutionary leap that combines the best of Exadata and the best of the cloud. Exascale reimagines how compute and storage resources are managed on Exadata platforms by decoupling and simplifying storage management. It ensures industry-leading database performance, availability, and security standards that organizations expect from Exadata. Coupled with the extreme performance of Exadata X11M, Exascale further increases operational efficiency with greater database deployment flexibility and simplified management.
Exascale features a reimagined approach to database snapshots and clones on Exadata. It enables space-efficient thin clones from any read/write database or pluggable database, significantly boosting developer productivity. Exascale seamlessly integrates with development, test, and deployment pipeline requirements while providing native Exadata X11M performance for all Oracle 23ai databases.
Databases on Exascale are automatically distributed across all available storage in the Exadata X11M storage servers, providing ultra-low 14-microsecond latency RDMA for I/O and database-aware intelligent Smart Scan with up to 500 GB/s throughput from XRMEM and flash and up to thousands of cores available to all workloads. Automatic data replication across multiple storage servers ensures fault tolerance and reliability.
Choice and Flexibility
Exadata X11M is available as an on-premises system, as Exadata Cloud Infrastructure in OCI, will be available through our multicloud partnerships, and as Exadata Cloud@Customer so customers can run Oracle Database cloud services in their data centers. In dedicated public cloud and hybrid cloud environments, customers can choose to run a mix of Oracle Exadata Database Service and Oracle Autonomous Database. In all these environments, customers have access to the same fully compatible Oracle Database and the same Exadata capabilities, enabling them to run their workloads wherever needed and without application changes.
In multicloud deployments, Oracle Database will run as OCI services on Exadata X11M in AWS, Google Cloud, and Microsoft Azure data centers. Customers can access all Oracle Database capabilities, including Oracle Real Application Clusters, to uniquely enable scaling and high availability during planned and unplanned downtime. In addition, customers can combine their applications, AI models, or analytics tools in AWS, Google Cloud, or Microsoft Azure with their data in Oracle databases supported by a low-latency network connection and take advantage of existing consumption commitments that they have to those cloud service providers.
In addition to the choice of where Exadata X11M is deployed and which capabilities to run on it, customers also have the flexibility to configure systems to match their current needs for database and storage servers and to expand them in the future incrementally. All deployment options can have as few as two database servers and three storage servers and grow to have dozens of each. In addition, on-premises customers have the flexibility to select different database server and storage server configurations.
Exadata Hardware
Exadata X11M Database Server
Exadata X11M database servers come with two of the latest 96-core AMD CPUs, which provide up to 25% faster per-core performance than the previous generation. Configurations deployed in OCI and multicloud environments come with 1,390 G of usable DDR5 memory per database server for all virtual machines; Exadata Cloud@Customer configurations offer 1,390 G, 2,090 G, and 2,800 G usable memory configurations per database server for all virtual machines. On-premises systems are available with 512 G of physical memory per database server in addition to 1.5 T, 2.25 T, and 3 T. Memory bandwidth in Exadata X11M database servers increases by 33% compared to Exadata X10M, greatly accelerating memory bandwidth-sensitive operations such as analytics and AI. Local storage on the database servers continues to utilize two 3.84 T NVMe SSD drives, which customers can expand to four 3.84 T drives.
Finally, Exadata X11M database servers continue to utilize the PCIe 5 data interface for the dual-port RDMA over Converged Ethernet (RoCE) network used for the private network.
Exadata X11M High Capacity and Extreme Flash Storage Servers
Each Exadata X11M High Capacity (HC) Storage Server (available for all deployments) and Extreme Flash (EF) Storage Server (available on-premises) uses two 32-core AMD CPUs, which are up to 11% faster than the previous generation.
The memory in the storage servers continues to utilize DDR5 and gets the same 33% memory bandwidth increase as the database servers, jumping from 4800 MT/s to 6400 MT/s. The higher memory performance provides a 21% improvement to Exadata RDMA Memory (XRMEM) read latency, improving from 17 microseconds in X10M to a staggeringly fast 14 microseconds in X11M!
Flash throughput increases up to 2.2x in Exadata X11M HC and EF storage servers as the four 6.8 T performance-optimized flash cards used for Exadata Smart Flash Cache move to PCIe 5. On Exadata X11M, Smart Scan throughput on HC storage servers is now 100 GB/s, up from 45 GB/s per storage server on X10M. In addition, Smart Scan can also now intelligently read from both XRMEM and flash, providing up to 500 GB/s of throughput per X11M HC or EF storage server or 11X what was previously available from just flash. In fact, a fully configured cloud system now has the potential to scan data at more than 31 TB/s – a truly staggering speed.
In storage servers, the RoCE Network Fabric utilizes the same dual-port RoCE network cards used in the database server.
The capacity of the 12 hard disk drives in the HC storage servers continues at 22 T each, for a total raw capacity per storage server of 264 T.
The four capacity-optimized flash drives in the EF storage servers also continue at 30.72 T for a raw storage capacity of 122.88 T per server.
Database Server X11M-Z
The Database Server X11M-Z is a new option for on-premises Exadata X11M deployments. It uses a single AMD EPYC processor with 32 cores and is available in 768 G or 1.25 T DDR5 Memory configurations. All other components in this server are the same as those in the standard database server discussed above.
High Capacity X11M-Z Storage Server
Also new for on-premises Exadata X11M deployments is the High Capacity X11M-Z Storage Server (HC-Z). It uses all the same components as the standard High Capacity storage server, with 6 x 22 drives, 2 x 6.8 T performance-optimized flash cards, and 768 G DDR5 DRAM (576 G for XRMEM). It provides the same ultra-low 14 microseconds 8K OLTP read latency as the standard HC storage servers, 50 GB/s Smart Scan throughput from flash, and up to 250 GB/s when scanning data cached in XRMEM.
Exadata X11M Database Server X11M-Z and High Capacity X11M-Z Storage Server. X11M-Z servers can be incorporated into existing or as the foundation of new Exadata Database Machine deployments, enabling even more flexibility for on-premises customers.
Exadata X11M Storage Expansion Rack
For on-premises systems, the Exadata Storage Expansion Rack X11M starting configurations support the latest Exadata X11M High-Capacity and Extreme-Flash Storage Servers. Customers can add more High-Capacity, Extreme-Flash, and High-Capacity X11M-Z Storage Servers to existing X8M, X9M, X10M, and X11M Storage Expansion Racks, ensuring they can scale their storage needs to meet the most demanding database deployments.
Exadata Software
Exadata System Software 25.1 was announced and released in December 2024. It includes support for X11M hardware and many new software features, including optimizations to AI Smart Scan, Exadata Exascale storage management, and faster ASM rebalance operations to increase performance, availability, operational efficiency, and security.
Accelerated AI Smart Scan
In Exadata System Software 25.1, AI Smart Scan has further optimized the top-K calculation so that each storage server maintains an Adaptive top-K list to reduce further the amount of data sent to the database from the storage servers. Adaptively calculating and tracking the top-K for a query at the storage server filters irrelevant data at the storage layer, reducing the amount of data returned to the database by up to 4.7X.
In addition to Adaptive top-K Filtering, Vector Distance Projection provides up to 4.6X faster Vector Search queries and up to 24X less data sent to the database. It avoids expensive recalculation of vector distances in the database server, freeing CPU resources for other work.
Finally, AI Smart Scan further accelerates AI query performance on vectors with INT8 and BINARY dimension formats by up to 8X and 32X, respectively, while maintaining accuracy above 92% compared to FLOAT32.
Exadata Exascale
Exascale can reduce the free space required to successfully rebalance data in the event of a failed disk drive or capacity-optimized flash device, from as high as 29% to as low as 3%. This is a significant improvement for machine administrators who must balance capacity, availability, and cost when sizing their Exadata platforms.
Automatic Acceleration of ASM Rebalance
Exadata X11M with Exadata System Software 25.1, and in combination with 23ai Grid Infrastructure, continuously monitors database I/Os and automatically adjusts the ASM rebalance power higher or lower to maximize performance and, therefore, reduce the time taken to rebalance extents without impacting database performance.
Extreme Performance and Efficiency, Available Everywhere
Exadata X11M is an extraordinary new generation of Exadata that enables customers to achieve extreme performance for all workloads—AI, mission-critical OLTP, and analytics—at the same price as X10M.
Extreme performance enables customers to run more workloads, with more data and users, on fewer systems. Coupled with intelligent power consumption optimization capabilities and the choice of deployment and operational model, Exadata X11M delivers extreme efficiency and saves customers money, power, and time.
Exadata X11M is available on-premises, as Exadata Cloud@Customer, in OCI, and will be available in Azure, Google Cloud, and AWS through our multicloud partnerships. This unprecedented level of choice lets customers develop their database applications anywhere and deploy them wherever they choose with the same 100% compatible Oracle Database and the same Exadata capabilities.
Alex Blyth is a Product Manager for Oracle Exadata with over 25 years of IT experience mainly focused on Oracle Database, Engineered Systems, manageability tools such as Enterprise Manager and most recently Cloud. Prior to joining the product management team, Alex was a member of the Australia/New Zealand Oracle Presales community and before that a customer of Oracle's at a Financial Services organisation.
Christian Craft
Senior Director of Product Management
Chris Craft is Senior Director of Product Management, focused on advanced Oracle database technologies including Oracle Converged Database, Exadata, Exadata Cloud Services, Autonomous Database, and Oracle’s Zero Data Loss Recovery Appliance. Mr. Craft has served in multiple roles within Oracle including On-Site Customer Support, Service Delivery Management, Sales Consulting, and Product Management. His experience ranges from application development, data modeling, and performance benchmarking to managing customer relations, and leading teams of technical professionals. Mr. Craft has worked with over 500 Oracle customers across virtually all industries and including many of the largest customers in the world.
Ashish Ray
Vice President
Ashish Ray is Vice President of Product Management within Oracle’s core database development organization. His product responsibilities include Exadata, Zero Data Loss Recovery Appliance, Oracle Database’s High Availability solutions (Maximum Availability Architecture), and all associated Platform as a Service (PaaS) Cloud Services.
