Next-Gen Power Solutions for Hyperscale Data Centers

The latest Cisco Cloud Index report shows us just how much evolution is happening within cloud and the data center platform.

Bill Kleyman

November 17, 2017

5 Min Read
data center racks

Today’s modern IT environment is rapidly expanding beyond the capabilities of their existing infrastructure. There are more users, a lot more data and new technologies that are carrying information over vast, widely-distributed distances. In fact, the latest Cisco Cloud Index report shows us just how much evolution is happening within cloud and the data center platform:

·       Traffic within hyperscale data centers will quintuple by 2020.

·       Hyperscale data centers will account for 53% of all data center traffic by 2020.

Furthermore, according to IDC, service providers will continue to break new ground in search of both performance gains and cost reductions as they expand their cloud architecture implementations. Additionally, the hosting-as-a-service model will continue to transition away from traditional models toward cloud-based delivery mechanisms like infrastructure as a service, spurring hyperscale growth in servers used for hosting (15% to 20% CAGR from 2013 to 2018).

The bottom line is this: In today’s digital economy, data centers must operate with hyperscale capabilities to meet demand, stay competitive, and provide new digital services. 

Hyperscale data centers are designed to be massively scalable compute architectures. To accomplish this level of scale and density, hyperscale data centers create optimizations around server utilization, energy efficiency, cooling, and their space footprint. One way they do this is by automating the delivery of critical resources. This ranges all the way from servers down to the racks.

However, how do you deploy hyperscale data centers at hyperscale speeds? If you’re going through a hyperscale deployment or a major data center refresh, it’s critical for you to think about rapid rack deployment, power efficiency, and how to get your data center up and running quickly. 

The Need for Hyperscale Efficiency

A recent report titled “United States Data Center Energy Usage Report” indicates that energy efficiency improvements will have saved 620 billion kWh between 2010 and 2020. The researchers expect total US data center energy consumption to grow by 4 percent between now and 2020. They predict the same growth rate over the next five years as it was over the last five years, reaching about 73 billion kWh.

Today, hyperscale data centers seek efficiency to lower costs. These new levels of efficiency lets you restructure cooling requirements, lower energy bills, and it allows you to put more compute in for the same amount of cooling. But, the conversation doesn’t stop there. Hyperscale capabilities also revolve around scalability and speed. So, how do you deploy hyperscale-ready data center faster and much more effectively? How do you deploy those critical racks faster to ensure optimal operation and reduce manual deployment errors? Similarly, what if you’re in a hyperscale data center today and are looking to do a big refresh? A lot of times your environment may already be capped when it comes to power consumption. You’ll have to find room to find room to improve and increase power efficiency and overall density. But, where does that ultimately come from?

New Technologies are Impacting Energy and Power Efficiency 

To help hyperscale data center deploy racks faster, with fewer errors, and a lot more efficiency, Server Technology recently implemented a feature that allows for the automated configuration of PDUs.

Basically, Zero Touch Provisioning (ZTP) is implemented via specific DHCP server options and Server Technology’s Configuration (STIC) protocol in conjunction with a simple TFTP server configuration pull. This process allows PDUs to be provisioned during the initial boot up, or whenever needed, to automate network setup, user permission updates, or other changes to the PDU configuration as needed.

ZTP allows for very fast deployment, one key to getting your networked PDUs up and running quickly. As soon as these units are up, they immediately begin to provide critical power and control information, as well as environmental monitoring within your data center. Remember, the ultimate reward for utilizing intelligent PDUs is the ability to gather valuable data points, monitor and control power, and make better decisions around capacity. 

ZTP is well-suited to hyperscale environments. These types of systems help:

·       Reduce time and cost for deployments

·       Automate processes to help minimize configuration errors

·       Eliminate the need for extra equipment

·       Bring improved scalability and energy efficiency

·       Allows administrators to deploy racks at rapid speeds for hyperscale requirements


Here’s the other cool part – you’re coupling ZTP systems with next-generation High Density Outlet Technology (HDOT). So, not only are you automating the delivery of PDUs into hyperscale data centers, you’re also leveraging powerful density capabilities with outlet technologies. High power densities and reduced cabinet space require new and innovative PDU systems. With Server Technology’s new outlet system, you provide industry standard C13 and C19 outlets in a drastically reduced footprint. In planning for future designs, HDOT helps combat the limited physical space that PDUs compete for in the data center rack. Solutions like those from Server Technology now leverage High Density Outlet Technology (HDOT).

­Final Thoughts

Within a hyperscale ecosystem, gathering critical pieces of information is key to success. With networked PDU deployments, you can integrate automation and data intelligence with your management architecture. This is why, when creating hyperscale capabilities, it’s crucial that organizations choose rack PDUs with remote monitoring and management capabilities. These types of solutions help deliver more data center intelligence, greatly improve power provisioning, and help you with capacity planning and remote control.

Hyperscale data centers are built with scale in mind and the ability to meet the needs of emerging use-cases. With new demands like big data, artificial intelligence, and other cloud applications all impacting how hyperscale organizations leverage their critical resources, hyperscale data center solutions will need to make scale-out and optimization easier. This translates to leveraging technologies that can help manage your most critical data center components. Hyperscale capabilities will continue to revolve around the intelligent delivery of resources while still improving efficiency. Managing power and automating the delivery of PDUs, for example, is a great way to design high-density hyperscale solutions.


About the Author(s)

Bill Kleyman

Bill Kleyman has more than 15 years of experience in enterprise technology. He also enjoys writing, blogging, and educating colleagues about tech. His published and referenced work can be found on Data Center Knowledge, AFCOM, ITPro Today, InformationWeek, NetworkComputing, TechTarget, DarkReading, Forbes, CBS Interactive, Slashdot, and more.

Subscribe to the Data Center Knowledge Newsletter
Get analysis and expert insight on the latest in data center business and technology delivered to your inbox daily.

You May Also Like