Modern Developer | November 6, 2017

High-Performance X7 Platform Generally Available

By: Karan Batta | Vice President, Product


During Oracle OpenWorld, we announced new compute instances based on our new "X7" hardware platform. These cutting-edge instances deliver on our vision to provide uncompromising performance and customer flexibility. Today, we're excited to announce that these new instances are generally available in our US East - Ashburn region, and will be available in other US and European regions soon.

New compute instances are available as bare metal servers, virtual machines, and as a new GPU instance. All three offerings are powered by Intel's latest Xeon Processors based on the Skylake Architecture enabling a host of new capabilities using new feature sets such as AVX512 vector instructions. Additionally, these instances provide dual 25Gb network interfaces for high bandwidth requirements. Oracle Cloud Infrastructure now offers the largest instances available in a public cloud today, with up to 52 physical cores per instance. 

The Standard VM and bare metal shapes are perfect for traditional high-performance applications along with computationally intensive parallel workloads such as CFD (computational fluid dynamics) and Monte-Carlo simulations.

The Dense IO shapes, also available as VMs or bare metal compute instances, support the most data-intensive requirements such as high-performance database applications and big data workloads. These shapes are backed by high-performance local NVMe drives providing consistent and predictable performance. 

Finally, GPU shapes powered by NVIDIA's Tesla P100 GPUs based on the Pascal architecture enable a new set of use-cases, such as Deep Learning Training and other modern enterprise workloads in various industries that utilize the CUDA platform. The ability to launch bare metal instances with GPUs is vital to workloads where even single digit percentage results in tasks potentially taking hours longer. 

We're also streamlining the experience for GPU instance users by providing images that are preconfigured with essential components such as the CUDA toolkit and the cuDNN library for seamless set up of AI workloads. At launch, we'll provide Ubuntu 16.04 LTS and Oracle Linux.

Launching these new instances is the same simple process you use today. You can launch them via our web console or use our developer friendly SDKs/APIs. For additional information please visit our Getting Started documentation which includes detailed instructions on launching, running and managing compute instances.

In addition, you'll be able to use your existing workflows to manage these instances in Oracle Cloud Infrastructure. For the instance lifecycle, you can leverage our Terraform provider with a single-click deployment, which makes it extremely easy to scale up and tear down resources as you need them. Most customers today are building HPC clusters with the intention of deploying containerized applications to make workloads portable. Hence, we've released a set of modules and an example base configuration that provides a Terraform-based Kubernetes installation to run a highly available and configured Kubernetes cluster on Oracle Cloud Infrastructure. 

See Us During Supercomputing Conference

Finally, we're going to be showing off all these fantastic new capabilities at the Supercomputing Conference from November 14 to 16 in Denver, Colorado. Our teams will be there in full force to showcase Oracle Cloud Infrastructure and what makes it so compelling for HPC, Enterprise Lift and Shift, and Cloud Native workloads. Our performance experts will also be speaking at NVIDIA's booth on Wednesday, November 15 at 11 a.,m., giving an overview of OCI and showing off our new X7 GPU instances.

If you'd like to talk to us or get started with some free credits, come see us at booth #1101 during the conference.

Join Us

We're always working hard to make our platform the highest performance, most flexible, and most customer-friendly. We love hearing from our customers on what's working and what needs improvement. Feel free to contact usWe're innovating at an even faster pace than before and we're always on the lookout for incredibly smart engineers and product managers to build the world's best cloud for enterprise customers. Drop us a line if you're up to the challenge!

Safe Harbor 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, and timing of any features or functionality described for Oracle's products remains at the sole discretion of Oracle Corporation.


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. Such statements are based on Oracle’s current expectations and assumptions, some of which are beyond Oracle’s control. All information in this article is current as of Novewmber 6, 2017 and Oracle undertakes no duty to update any statement in light of new information or future events.

Vice President, Product

Karan is runs product management at Oracle Cloud Infrastructure for Hardware across OCI and HPC. He leads vision and product roadmap for compute infrastructure components and hardware. He regularly speaks at conferences and user-groups worldwide such as GPU Technology Conference, Super Computing & Siggraph. Previously Karan was leading accelerated hardware and infrastructure development for major public cloud providers and successful startups.

More about Karan Batta
This is a syndicated post, view the original post here