The Role of an Adaptive Data Center in an Anytime, Anywhere World

In this week’s edition of Voices of the Industry, Jakob Carnemark, CTO, CEO at Aligned Energy, explores how increasing and dynamic workloads require an adaptive data center. 

The effects of an anytime, anywhere world are felt across industries. Retail, healthcare, agriculture, transportation, manufacturing, energy. No sector goes untouched by the growth of digital business.

JAKOB CARNEMARK, Aligned Energy

Whether it’s B2B or B2C, customers expect more from the goods and services they buy and the businesses they buy from. They want every business to be responsive and creative in meeting their needs. And over the past 20 years, businesses have responded. They have responded, in large part, with technology. IDC calls digital transformation “the biggest industry shakeout since the Industrial Revolution.”

Yet, data centers – the infrastructure that supports the digital world – have been slower to embrace change.

Perhaps the greatest impact of our anytime, anywhere world is happening now in the data center. Many trends – from the rising tide of digital business, IoT, and virtual reality to the increasing value of data analytics and the rise of high performance computing – dramatically increase the variability of workloads and the density demanded of the data center. To support those trends, the data center must become more adaptive.

Growth is matched by increasingly dynamic compute loads

As digital business grows, data centers are asked to support exponentially increasing compute loads (high density) that are increasingly dynamic (variable density).

By 2020 annual global Internet traffic (a reasonable proxy for compute load trends) will reach 2.3 zettabytes, according to Cisco. Global Internet traffic in 2020 will be equivalent to 95 times the volume of the entire global Internet in 2005. To accommodate this huge increase in workloads, IT infrastructure (servers for compute and storage, and networking equipment) has evolved; for example, servers can now run at much higher densities (as high as 50 kW).

At the same time as compute loads are growing exponentially, they are becoming more dynamic as well as demand varies from month-to-month, day-to-day, hour-to-hour, and project-to-project. An ecommerce site’s traffic spikes at the end of November through the end of December, for example. Traffic to a search engine’s U.S. data centers is high during the day and lower in the middle of the night, though then traffic to its Asia data centers is high – demand for compute, in this case, “follows the sun.” Compute workloads also depend on the business cycle – development work might be very high for a period, then drop off, then spike again. Again, IT infrastructure has evolved to accommodate. For example, servers support increasingly variable workloads, using less power when utilization is lower and ramping up when necessary.

IT infrastructure has kept pace, but data center infrastructure has not

IT infrastructure has kept pace with the amount and dynamic nature of the work it’s asked to do. But innovation in software and hardware has dramatically outpaced innovation in the foundation of it all – the data center. With the exception of hyperscale data centers, the data center infrastructure that supports compute loads hasn’t changed much at all.

Servers can run much hotter today than in the past, the burden for heat removal falls on the data center. Most data centers can support high density servers, but the only way to ensure that they don’t overheat is to either half-fill the racks or spread them apart. Both approaches create stranded capacity – a significant inefficiency, and a key driver of low infrastructure utilization rates. Indeed, typical enterprise infrastructure utilization rates are 30-40%.

All about the infrastructure

A data center’s ability to support high density or ramp up and down in response to workload demand is driven by its how dynamic its infrastructure is. Traditional approaches to power and cooling have evolved around relatively stable loads of similar density. That is not the compute of today. Cooling systems need to be efficient and stable under dramatically varying loads and delta T’s. The Aligned Energy’s cooling platform has a consistent energy overhead from 10%-100% load and is able to respond to highly variable delta and air requirements.

The electrical grid today is dramatically more dynamic with the increased introduction of renewable energy. This coupled with more dynamic and growing data center loads will continue to put stress on the grid. To address these issues, the grid is moving toward frequency control as a primary driver for reliability. Frequency changes when grid supply and demand are unbalanced. Aligned Energy has been developing and deploying the foundational technologies to allow data centers to be dynamically adjust for and protect against frequency imbalance.

Increasing and dynamic workloads require an adaptive data center

How do IT operations and infrastructure leaders enable the business to respond to consumer demands with innovation at every layer of the IT stack – from the software to the data center? It requires the data center to become for IT exactly what IT has had to become for the business: adaptive. At Aligned Energy, the adaptive data center is a function of our intelligent infrastructure and a culture of innovation.

This world eaten by software, where workloads follow the sun and business models are incredibly dynamic, needs data centers that are able to handle dynamic power draws and dynamic densities. So what we’ve done from the cooling perspective and the electrical perspective is architected traditionally static systems to be dynamic, able to handle different densities as well as variable power draws (where traditional data center infrastructure is static and doesn’t like highly variable workloads).

For example, our patented cooling system can handle much higher delta Ts than is typical because it removes heat at its source, with heat sinks close-coupled with the server racks. We support servers running much higher densities and just as efficiently support servers running at very low densities – from 1 kW per rack up to 50 kW, within the same row – because our variable speed fans and variable frequency drive (VFD) pumps allow flow rate to respond to changing loads.

To be clear, the value of an adaptive data center is not in making the data center layer look more like the software layer. There are core economic drivers at work here: utilization rates of 30-40% represent a significant waste of capital and energy, and stymie a business’ attempt to nimbly respond to changing customer demands.

Bottom line

IT is no longer just “business support”; it’s the engine that drives digital business. For IT, supporting the adaptive business models required to meet customer demand has meant becoming adaptive, and that goes for the data center too. The goal: align data center supply (power, cooling, and space capacity) with IT demand – so that IT can align with the business to deliver whatever the customer wants next.

In this “biggest industry shakeout since the Industrial Revolution,” the companies that survive will be those that adapt. “All failure,” as Max McKeown says, “is failure to adapt. All success is successful adaptation.”

Jakob Carnemark is Founder, CTO and CEO at Aligned Energy, an infrastructure technology company that offers colocation and build-to-scale data center solutions to cloud, enterprise, and service providers. Our intelligent infrastructure allows us to deliver data centers as a utility—accessible and consumable in the amount and moment needed. By reducing the energy, water, and space needed to operate, our technology innovations offer businesses a competitive advantage by improving reliability and their bottom-line, while helping secure the health of the planet. Learn more and get in touch via alignedenergy.com.