Virtual Interconnection: New Technology For Cutting Edge Data Centers
Virtual interconnection is a new frontier that’s seeing rapid acceptance by colocation and cloud service providers. In a new DCF report, DataBank shares its take on the near-term future of software-driven IT infrastructure.
Last week, we continued our special report series on virtual interconnection. This week, we’ll look at the benefits of virtualizing interconnection.
Interconnection is a core service of most co-location providers that supports many-to-many connections between entities for rapid data exchange as well as an on-ramp to cloud services. The need for interconnection is growing in lockstep with the move toward digital business as constant connectivity becomes an essential element of partner, supplier, and customer relationships. It is the fastest-growing segment of the colocation industry, with worldwide bandwidth more than quadrupling between 2017 and 2020 to 5,000 terabits per second.
Interconnection virtualization is a relatively new technology frontier that is rapidly gaining acceptance by colocation and cloud service providers. It permits businesses to create virtual cross-connects between customers, workloads, clouds, data sources, and applications, allowing data center workloads to become more distributed and flexible while also distributing processing and data more economically and efficiently. Slow, manual provisioning gives way to instant and programmable management.
IT managers can self-administer infrastructure instead of paying carriers to do it. They can also better adjust and align infrastructure with the demands of their applications in real-time, reducing the need to overbuild and leave expensive equipment running idle.
Interconnection virtualization is growing in response to several major trends sweeping the IT industry:
One is the as-a-service model that offers the appeal of flexible subscription-based pricing, scalable capacity and OpEx predictability. IDC estimates that 75% of enterprises will use some form of as-a-service consumption this year, led by a three-fold increase in demand for on-premises infrastructure delivered on a pay-as-you-go basis.
This rapidly growing technology delivery model requires agile network infrastructures that can adapt to sometimes unpredictable changes in network traffic flows. As business operations become more dependent upon services hosted in the cloud, resilient and scalable networks will be table stakes for businesses.
The second driving factor is edge computing, an evolution of cloud infrastructure that moves data collection, processing, and services closer to the point of value. The market for edge products and services is poised for hypergrowth, with expansion expected to proceed at 37% annually through 2027, according to Grand View Research.
There is a common myth that edge computing is all about managing small devices. While that so-called “far edge” component is an important part of the market, the definition of edge computing is more expansive. “Near edge” and “middle edge” architectures are typically built in the vicinity of data centers in tier 1 and tier 2 markets to serve large populations of users or to support latency- sensitive compute requirements. These regional hubs support applications that can’t practically be served by central clouds.
Edge environments can span multiple servers and even encompass small data centers. Over time, those facilities can become quite large as new applications and user populations are added. Edge computing
also encompasses data center infrastructure that is provisioned to meet regulatory requirements for data locality or data sovereignty as well as application- specific uses such as support for retail outlets.
It is impractical for regional cloud data centers to provide the sub-second response times that many new applications require. Edge computing will support latency-sensitive applications like augmented reality, personalized video streams, remote vehicle control, and real-time recommendation engines that will fuel future business growth.
A third trend driving the growth of interconnection virtualization is the rapid deployment of 5G networks. One billion 5G subscriptions are expected to be in use worldwide by 2022 and an estimated half of all mobile connections could be 5G-enabled by 2023. The high-band mmWave service that delivers wireline speeds to untethered devices is expected to support entire new categories of applications:
- Factory and warehouse robots will be able to roam more freely without the constraints of network cables. Real-time data streams will let manufacturers and utility companies apply predictive maintenance to repair and replace equipment in the field before it fails.
- Advanced transportation infrastructure such as sensor-equipped traffic controls, smart lighting, and smart meters will make urban environments safer and more efficient.
- Drones equipped with image recognition capabilities will take over inspection tasks that are too dangerous or impractical for humans to perform.
- Sports teams and stadium owners will deliver video feeds customized to each fan’s preferences.
- The quality and sophistication of remote health delivery like robotic surgeries and diagnoses will expand.
- Immersive classroom experiences will give each student a personalized journey down the Amazon River or the Martian landscape.
5G services will also change the way network topology is designed. Data-intensive applications will require more compute power to be located at the edge. New sites will be needed that operate at shorter distances and higher speeds.
Download the full report, How Virtual Interconnection Supports Distributed Digital Business, courtesy of DataBank to learn more. In our next article, we’ll explore why data centers must rethink interconnection. Catch up on the last articles here and here.