Oracle Cloud infrastructure delivers Industry’s highest levels of performance for production enterprise applications
Oracle OpenWorld, San Francisco, Calif., Oct 3, 2017 -- Oracle today announced enhancements to Oracle Cloud Infrastructure that provide enterprises with the highest levels of performance and reliability delivered by any major cloud vendor. These enhancements to Oracle Cloud Infrastructure Compute, Storage, Networking, and Edge offer IT operations, developers, and researchers, infrastructure services optimized for production enterprise applications.
Achieving high performance and reliability for enterprise applications and production workloads is capital and resource intensive for most customers in their own datacenters and even more difficult in the public cloud. Oracle Cloud Infrastructure delivers a reliable foundation for production enterprise applications, providing unique enterprise reliability features and improving compute performance by 87 percent, with 54 percent better price-performance[i] compared to other major cloud providers. Now, back-office applications like enterprise resource planning (ERP), public-facing database applications like e-commerce, and performance-intensive applications like high-performance computing (HPC) run faster with greater reliability and at lower cost on Oracle Cloud Infrastructure.
“Enterprises are moving workloads to the cloud for a variety of reasons, whether it is to eliminate datacenters, migrate legacy workloads, or build and run performance-intensive applications in a more agile environment,” said Don Johnson, senior vice president, Oracle. “Running traditional and cloud-native production applications successfully in the cloud requires consistent high performance and reliability across the stack. Keeping all of these factors in mind, we have continued to improve Oracle Cloud Infrastructure to deliver industry-leading performance with simple, easy-to-understand pricing, making Oracle the most cost-effective cloud to run serious enterprise applications.”
Oracle enables enterprise reliability by being the first vendor to support trusted technologies such as local database clustering through Oracle Real Application Clusters, and Oracle Exadata as a managed cloud service. When combined with geographically distributed datacenter regions, each with multiple fault-independent availability domains, Oracle is providing a unique set of capabilities to maximize application availability and reliability.
Enhanced performance is made possible through a non-oversubscribed 25 Gigabit (Gb) network and optimized edge access with an industry-leading DNS. In addition, Oracle provides the best price and performance per CPU and GPU core, guaranteed storage performance at no additional cost, the lowest cloud data egress and access costs, and the most flexible pricing mechanisms to bring customers a better experience at the lowest prices in the industry.
Industry-Leading Compute to Boost Performance
In order to bring enterprises improved performance, Oracle has introduced an industry-leading system architecture (X7) and a new line of compute instances. The new Bare Metal Dense I/O instance provides 52 cores and 51 Terabytes (TB) of non-volatile memory express (NVMe) solid state drive (SSD) capacity, capable of millions of input/output operations per second (IOPS). These bare metal instances will improve performance for workloads related to simulations, computational fluid dynamics, analytics, and high scale enterprise applications. The new Bare Metal Standard instance offers the same number of cores, but can scale to 512 TB of SSD block volumes, enabling high storage capacity use cases no other cloud vendor can match. New Virtual Machine instances will also be available in 1, 2, 4, 8, 16, 24-core shapes, leveraging the faster processors and increased memory of the X7 architecture.
Oracle’s new bare metal GPU instances, which are up to 400 percent faster while costing 58 percent less than comparable solutions from competitors[ii], enable high performance for compute-intensive workloads used by researchers, analysts, developers, and animation professionals. GPUs are designed to process and analyze massive data sets more efficiently than general-purpose processors, making it ideal for complex machine learning and artificial intelligence algorithms. These compute-intensive workloads are increasingly being run in the cloud due to the ability to deploy vast resources, obtain results, and release the resources more quickly than any other deployment model. This capability provides the ability to pay for only what a user needs to complete the desired computational task.
High Performance Storage with Guaranteed Performance at No Additional Cost
Oracle is the first major cloud provider to guarantee storage performance at no additional cost. Oracle’s new all-flash block volume storage is specifically designed to deliver within 10 percent of 60 IOPS per Gigabyte (GB) and 480 Kilobytes per second of throughput, 99.95 percent of the time.[iii] Optimized for performance-intensive, production enterprise workloads, Oracle Cloud Infrastructure customers can now run resource-intensive workloads, such as databases and analytics, with block storage that is 25 percent faster and costs up to 98 percent less than the competition.[iv]
Oracle is also introducing a new Archive Storage service, leveraging the same APIs as the Oracle Cloud Infrastructure Object Storage, available at only $0.003/GB/month. This offers customers a low-cost, integrated alternative for long term storage.
Enhanced Networking and DNS for High-Bandwidth Workloads
As companies increasingly utilize high-bandwidth applications and interact with customers in near real-time, applications must respond quickly with minimal latency. Oracle’s new 25Gb network infrastructure offers significantly more throughput than most traditional cloud network architectures and allows enterprises to cost effectively take full advantage of the latest connectivity to compute, storage, and database services.
Integrated DNS from Dyn, the leading cloud-based internet performance and DNS provider, is also now available on Oracle Cloud Infrastructure. This best-in-class edge solution enables customers to monitor, control, and optimize enterprise back-office and public-facing applications to deliver faster access and greater end-user satisfaction.
Platform Enhancements and Integrated Solutions for Easy Migration
Oracle Cloud Infrastructure offers packaged migrations for Oracle E-Business Suite, JD Edwards, and PeopleSoft, helping customers instantly modernize their infrastructure on the best price-performance cloud. Oracle is also making a new data transfer appliance available to import large quantities of data into the Oracle Cloud, as well as a new tagging service to support the management, monitoring, and automation of large amounts of resources.
Oracle Cloud Infrastructure also announced a new version of its Ravello Service, which offers up to 14X better performance than the previous version. Enterprises can now run VMware VMs of up to 32 vCPUs to scale applications vertically and up to 2,000 VMs to scale horizontally. This increased performance and scale is combined with new high availability capabilities to enable customers to migrate applications with higher reliability requirements.
Oracle has also expanded its cloud infrastructure ecosystem with 25 certified tier-1 system integrators (SI) and its new Oracle Cloud Jump Start program enables customers to evaluate solutions offered by Oracle’s ISV and SI partners on Oracle Cloud Infrastructure at no cost. The Oracle Cloud Jump Start program is launching with solutions from partners, including Apprenda, App Associates, Astute, DataStax, Denovo, Forsythe, Mythics, Sunera, and Qubole.
[i] Based on industry-standard SPECint_rate2006 benchmarking, a measure of multicore, multisocket workloads. Compared to AWS M4.16xlarge
[ii] Based on a standard benchmark with TensorFlow as the main Neural Net Training Framework, AlexNet and Inception V3 computer vision models, and the ImageNet dataset. Compared to AWS p2.8xlarge and Azure NC24
[iii] Up to a maximum of 25,000 IOPS and 320 Megabytes per second throughput per volume
[iv] Based on a typical enterprise workload of 400 Gigabytes and 20,000 IOPS. Compared to Amazon EBS Provisioned IOPS SSD (io1) volumes
|DSS Home | About Us | Contact Us | Site Index | Subscribe | What's New|
|Copyright © 1995-2015 by D. J. Power (see his home page). DSSResources.COMsm is maintained by Daniel J. Power. Please contact him at djpower1950 at gmail.com with questions. See disclaimer and privacy statement.|