A Quick Guide to Choosing the Right Data Center Design

Digital transformation is driving modern businesses to grow-up fast. According to research by Innosight, the average age of companies on the S&P 500 is expected to shrink from 33 years in 1964 to just 12 years by 2027.[i] At the same time, data center design is evolving to serve the current needs of enterprise IT. But similar to choosing a home, the criteria you base your decisions on can change over time. The requirements you had when settling on that first one-bedroom apartment probably no longer serve the needs of your growing family. Likewise, knowing your long-term utilization needs – and understanding a little about data center design – is essential for making the best decisions when it comes to your current and future IT architecture.

Data center design foundations

On the surface, many fundamental design principles and practices have remained consistent as data centers designs have evolved over the years. Facility owners, while focused on up-time, continually seek to optimize data center operations and incorporate best practices. The goals are known: deliver improved efficiency, resiliency, flexibility, security and value – we have heard all these before.

Location also matters. Having efficient and cost-effective operations means selecting geographic locations based on proximity to energy sources and inexpensive property, whereas providing the best performance and quality of service (QoS) means selecting locations near to dense concentrations of connectivity, businesses and users.

But today, virtualization and other emerging technologies are changing the way we approach IT, compelling the industry to re-think how we deploy purpose-built data center designs.  Here are a few things you need to know about data centers and data center designs as you architect for the future.

On-premises data centers could be a single rack in a server closet or a full facility built by a company to serve its specific business needs. The operational cost (and risk!) is born by the parent company, driving businesses to focus on “getting out” of on-premises data centers. Ideally, businesses could move everything to the cloud to mitigate risk and minimize cost, but the reality is that some data is too sensitive and some workloads too critical to move off-premises. With additional constraints imposed by data regulations and governance requirements, there will always be a need for some amount of on-premises IT infrastructure.

When a company’s digital transformation needs outpace the capital resources and IT operating footprint, choosing a colocation facility as a base of operations can be an excellent option. Colocation facilities, simply put, house IT infrastructure on behalf of a customer. Approaches can vary in scale from as small as a half-cabinet to Retail, Wholesale or Hyperscale-sized deployments. But no matter the scope, there are common core design principles around building and operating these facilities to deliver the secure performance, efficiency, flexibility and resiliency that today’s IT teams expect.

Designing for the future, means designing for flexibility

Regardless of the intended use, data center design is moving toward a standardized, modular approach for big and small facilities alike. While custom builds can be great as you get exactly what you ask for, they also entail larger capital investments and longer provisioning times. Without standard design and operational practices, a custom approach can also lead to operational and maintenance risks. By contrast, a set of proven data center design principles that work together in a modular manner enable facility owners to quickly provision solutions that can flex with business needs on-demand.

But today, virtualization and other emerging technologies are changing the way we approach IT, compelling the industry to re-think how

1. Going big with Flexible Data Center (FDC) design: While sites and locations are all different, FDC uses a standard catalog of design elements that can be adapted to individual site requirements without compromising integrity. Common parts enable supply chain and capital cost efficiencies while yielding long-term maintenance predictability. For Equinix this means higher uptime/availability and significantly lower operating risk across people, parts and procedures.

Any space on the data center floor not dedicated to colocation is a lost opportunity – it’s not making money for the owner, and it’s a missed interconnection opportunity for the other tenants. Standard designs minimize the whitespace while ensuring that proven designs and best practices are used regardless of intended use, location or construction contractor. This approach avoids utilization-arbitrage in those empty spaces and maximizes scalability, predictability, and flexibility for every tenant deployment. Maintenance and other data center services can also easily be incorporated in programmatic blocks of density, while on-demand and planned services operations may continue without risk to neighboring tenants.

2. Going small with a modular Edge Data Center (EDC) design: The push to data-intensive applications such as the internet of things (IoT), the rise of 5G and connected vehicles means that the network and data solutions we have today must evolve or be left behind. In kind, the nature and role of the edge data center is evolving as well. Where previously large interconnected ‘hub’ data centers in major metros were considered to be the edge, today those same facilities are considered to be the regional core. The evolving edge is one that requires an even more fine-grained approach and a new set of very small, typically un-staffed facilities that can support these new data-and-traffic intensive distributed workflows. Key design principles for EDCs include:

• A consistent approach to manufacturing to dramatically reduce variability from site-to-site, delivering time-to-value for tenants and operating and maintenance efficiency.
• A modular design that can flex to meet localized business needs while reserving the ability to contract or expand as operations demand.
• Flexibility to scale/adapt common design without compromising integrity across an array of demanding physical shapes and locations.
• The ability to deploy capability where it matters for the application, including edge-proximity which commonly means densely-packed, space-constrained locations.
• An interconnection and connectivity-forward design that delivers against the data-intensive role the edge plays in demanding modern workloads.
• An energy solution that delivers resiliency, redundancy and a responsible approach to sustainable operations.

To learn more about the trends driving the multi-tenant data center and services industry, download the white paper.

You may also be interested in reading about our about our global data centers and Platform Equinix.

[i]Innosight, 2018 Corporate Longevity Forecast: Creative Destruction is Accelerating.