Thanks to a variety of new data-powered technologies it makes sense that modern data centers are working harder than ever. The harder they work, the more heat they put off and the more strain it all puts on the cooling system keeping temperatures moderated.
There’s another aspect to all that use of power, however. It’s incredibly expensive. Not only do the data servers and systems require loads more power, so does the climate control system. That’s why many innovative data center technologies explicitly deal with cooling.
Another consideration is the far-reaching impact on the environment. More power means more fossil fuels in use, higher emission levels, and faster deterioration for the environment. It’s all connected, like it or not.
But with the right optimizations, energy consumption can be mitigated, cooling can be more effective and the servers themselves can experience better performance and longevity.
Here are some of the more popular cooling trends currently making waves in the data center field:
Conventional data center cooling systems are designed to remove heat from inside the server space through an exhaust system and then introduce cooler air via the AC. Geothermal does something much different.
It relies on the temperature of the surrounding earth, creating a natural heat sink. These types of systems utilize subterranean geology to naturally cool a heated space, like a server room. The biggest benefit is that the system is much less costly, as well as being more environmentally friendly.
Several data centers already use this method, including the American College Testing data center in Iowa City. The 4,000 square-foot space is cooled by a geothermal “bore field” with a water-based piping system.
While closer to conventional cooling systems, P2P or pumped two-phase systems offer a much higher cooling efficiency than both air and water cooling at a fraction of the cost. In addition, energy consumption is greatly reduced because there is no HVAC unit or chiller involved.
Space savings are also offered because the entire system takes up much less space, leaving more room for bigger or additional servers and equipment.
Liquid flows through the isolated system across a cold plate which is itself directly attached to a heating unit. The heat vaporizes the liquid and is carried away while the vapor moves to a condenser. Inside the condenser, the vapor is returned to liquid form releasing the heat. Any excess liquid is then moved to an accumulator and then eventually rolled back into the system. The entire cyclical process offers incredibly effective cooling for a heated space.
This method tends to be effective only in certain regions or locations. In a hot humid climate, for example, natural air is going to cause problems in an already warm space. But in colder areas, the natural and surrounding outdoor temperatures can be utilized to cut down on cooling costs. The same is true of natural water flows.
The natural method is a little more bleeding edge than most and must be factored into the overall design of the data center. That means effectively making use of natural air and water flows when the entire facility is being developed. Still, it’s a great way to cut down on operational costs and make use of the surrounding environment.
One example of this is placing an operational data center underwater. The cool temperatures of the surrounding water body will keep the space much more controlled.
With most liquid cooling and water-based solutions, the liquid flows through surrounding pipes and coils to keep a system or facility controlled. Immersion cooling works the same way, except the computer systems and servers are specifically submerged in the liquid.
Either the system can be encased in a protective shell and then submerged, or simply doused in the liquid as-is. The problem with the latter scenario is that most fluids and liquid cooling solutions are caustic, which ruins the electronics. More recent developments are exploring the use of mineral oil as a cooling component which may be safer to dunk electronics in.
Finally, we have a slightly different trend that involves more effective use of the existing cooling systems. Deploying a more eco-friendly system alongside IoT would provide massive benefits in cost savings, but smarter monitoring can still improve the situation on its own.
Essentially, smarter IoT systems are designed to automate cooling inside areas with greater accuracy and reaction times. A smart thermostat, for example, will take an accurate temperature reading of a surrounding area or room and then power on the cooling system. Once the temperature reaches a suitable level the system is powered down again. Adversely, the system can also be used to make minor adjustments that help maintain the appropriate levels.
It’s not unlike how smart home thermostats work such as Nest which does exactly the same thing, only in a residential space.
Believe it or not, these solutions are becoming more numerous because they are so necessary. Cooling and climate control systems are the most costly element of running a data center, outside of maintaining and supplying the operational hardware. The systems must continue to operate indefinitely, and when there’s a problem it wreaks havoc on the entire operation. Servers can fail resulting in the loss of data, operations grind to a halt meaning uptime ratings take a hit, and all of it will surely result in the loss of customers.
When it comes to big data and cloud computing technologies — data centers serve as the underpinning — downtime is detrimental to the entire business model. So, the cooling not only needs to be reliable but effective, and for extremely long periods of time lest the business descends into chaos.
No one wants that, which means data center cooling solutions need to get bigger, more effective and much cheaper. The above trends are the answer and will afford data center operations higher capabilities.
Kayla Matthews writes about data centers and big data for several industry publications, including The Data Center Journal, Data Center Frontier and insideBIGDATA. To read more posts from Kayla, you can follower her personal tech blog at ProductivityBytes.com.