The success of server virtualization has opened the door to virtualizing other datacenter components, such as network and storage, which is leading the next wave of enterprise datacenter optimization — software-defined infrastructure — which seeks to virtualize the entire datacenter infrastructure as a new flexible hardware layer under the hypervisors, containers and native applications offering workload optimized hardware infrastructure on demand.
Accelerating the spread of this new architecture is an industry-standard technology blueprint developed by Intel called Intel® Rack Scale Architecture.
At Ericsson, we are leading hyperscale development with our Ericsson Hyperscale Datacenter System 8000, which is a modular system based on Intel Rack Scale Architecture. And we have already seen high demand for software-defined infrastructure from some of our top customers, including:
- SK Telecom – which will accelerate the realization of its software defined infrastructure vision
- Far EasTone – which will explore opportunities for enterprise hyperscale cloud
- Telstra – for which we are delivering a telecom cloud solution including Network Functions Virtualization (NFV) and software-defined networking (SDN)
- Telefónica – which chose Ericsson for its UNICA project, and aims to virtualize network functions across all Telefónica's operations. Ericsson will supply its new Hyperscale Datacenter System 8000, Ericsson Cloud Manager, Cloud SDN and automation, as well as systems integration and support services
- Swisscom – which will deploy an Ericsson full stack telecom cloud solution for its core network functions, including NFV infrastructure and software-defined networking
However, interest in software-defined infrastructure doesn’t stop with telecom operators. On the contrary, we see significant interest from major enterprises and municipal institutions outside the telecom sector, including both enterprises that operate their own datacenters and enterprises that use other companies’ datacenters.
Many of these companies work with thousands of applications, and they are looking for better utilization of resources and to become more agile whilst retaining the robustness and high degree of security that comes from a completely separated infrastructure. Research has found that companies implementing Ericsson’s hyperscale platform can capture significant TCO savings compared to traditional datacenter infrastructures — generating significant capex savings of up to 59 percent, opex savings of up to 75 percent, and a return on investment (ROI) of up to 149 percent for large enterprise datacenter operators over a five-year period.*
A key component of these TCO savings comes through a new, open and flexible manager that can also handle third party product (3PP) equipment. The Ericsson Command Center configures and manages compute resources, storage capacity, and network connectivity of both Ericsson Hyperscale Datacenter System 8000 and systems from other vendors. One of the most interesting features of our system is the ability to partition the infrastructure (one or multiple PODs) into vPODs – that are compositions of compute, storage, and network resources that provide a complete and isolated bare-metal infrastructure optimized for a specific workload. The workload experiences the vPOD as its own private POD, although it is flexible to its size and content.
The computational foundation of Ericsson's first generation of Ericsson Hyperscale Datacenter System 8000 is based on Intel's 4th generation Intel® Xeon® processor E5 product family, enabling best in class performance and functionalities necessary for a modern software defined datacenter implementation.
Taking a broader view, we see a great need for standardizing software-defined infrastructure. As my colleague Howard Wu, head of SDI at Ericsson, wrote in a recent post, “we want to drive innovation and openness in the datacenter infrastructure industry, and we believe this is one of the keys to accelerate digital industrialization for 5G and the IoT.”
With that in mind, we have joined both the Open Compute Project (OCP) as a Platinum member and Open Data Center Committee (ODDC). Both are great forums, in which we will work together with Intel to drive the adoption of optical interconnect, pooled systems and management interfaces for Intel Rack Scale Architecture compatible components.
Are you ready to learn more about how SDI can help lower TCO and improve agility? If so, start by exploring this report: An Economic Study of the Hyperscale Datacenter.
*For more information, visit: https://cloudblog.ericsson.com/how-to-increase-your-agility-while-reducing-datacenter-operations-costs