Intel has had a long-running history of accelerating data center transformation by simplifying complex technologies, establishing industry best practices and driving broad adoption. We are committed to enabling Cloud Computing across cloud service providers, comms services providers and enterprises. Because running a Cloud at scale isn’t easy, we drove the Cloud for All initiative with leading software vendors and actively contribute to groups like the Open Compute Project and the Open Data Center Alliance to lay the groundwork for standards-based next-generation solutions.
This week, we’re excited to share our latest contributions to the community at the OCP Summit 2016 in San Jose. One of the new innovations is the design of a next-generation data center storage architecture that pools high-speed NVM Express™ solid-state drives to accelerate storage performance. This design, codenamed “Lightning,” was co-developed by Intel and Facebook. It is designed around OCP rack specifications and available for broad community contribution and adoption.
Lightning is designed to scale across different storage workloads, including capacity storage, pooled storage, and high-availability storage. It delivers the benefits of NVMe, including fast storage performance and great scalability. These are critical for emerging workloads in the Cloud such as big data analytics and machine learning that demand superior IO performance and large capacity from the infrastructure.
Another important outcome of Intel and Facebook collaboration has been the development of our latest 16-Core Xeon D system on a chip (SoC), our highest performing datacenter SoC to date. This SoC was optimized for Facebook’s workload requirements and is enabled in an OCP form factor design.
Because Cloud customers at scale often have workloads that have an affinity for acceleration, we are advancing innovation in the arena of field-programmable gate array (FPGA) technology. FPGAs provide flexibility in the acceleration of an increasing variety of data center workloads, and our customers are requesting new classes of products that incorporate both FPGAs and CPUs.
At the OCP Summit, we announced plans to deliver libraries in combination with next generation Intel® Xeon® processors and Arria® 10 GX FPGAs on a multichip package for software development. These libraries can be downloaded to the FPGAs for programming their functionality. We plan to bring these libraries to the open source community, fueling innovation within the FPGA developer ecosystem in the form of new algorithms and applications for end users, and will speed time to market for accelerator-based solutions. We also plan to work with the OCP ecosystem to enable future OCP designs to accommodate the integrated Intel Xeon + FPGA product.
Finally, in another step towards open infrastructure solution enabling, Intel and partners are showcasing open source OCP hardware interoperability with Intel® Rack Scale Architecture. Intel Rack Scale architecture is a logical architecture that enables scalable pooled resources and higher levels of data center wide resource utilization. Intel Rack Scale Architecture Management APIs can be used to manage OCP compliant hardware such as servers and storage, and create composable resources that can be allocated to applications on demand based on their requirement. This combination creates a very simple ‘open’ path for software defined infrastructure (SDI). WiWynn and Quanta are showcasing early versions of their work, and Ericsson recently highlighted their intent, thereby building a strong ecosystem to drive this interoperability.
As a founding member of the Open Compute Project, our support continues to grow with 18 contributions to date and enablement influencing greater than 40 partner products. As we foresee the trends of emerging cloud workloads, we continue to invest in shaping the technologies required for our customers to build world class Cloud infrastructures through standards-based next-generation solutions.