Engineers from La Salle-URL share the latest news and projects in the field of network solutions in telematic engineering.

30 April 2019 | Posted by bernat.rovirosa

Green Datacenters

Since the beginning of computing, the increase of performance has been the main objective (faster processors, more memory, etc.) However, this objective can lead to resource waste, having a significant impact on the environment. A solution to this problem is the implementation of Green Datacenters.

These Green Datacenters, aside from not using certain resources which directly contaminate the environment (chemicals, etc.), are all about reducing energy consumption while maintaining performance. The cause is that nowadays, datacenter energy consumption almost makes up 2% of the worldwide energy generated, and this number is only predicted to increase.

That would not be a problem if all the generated energy was green, but that is not even close to the truth. According to 2017 data, of all energy, only 26.5% is renewable, and not all of it is green, which means that statistically, approximately 70% of datacenters contaminate by default.

To reduce environmental impact, here are some of the strategies that can be used in a datacenter:

  • Low-spec hardware: Not all hardware consumes equally. What is more, in some cases, the balance between energy consumption and processing power of lower-spec hardware is better than that of cutting-edge hardware, meaning that in higher quantities, energy efficiency with low-spec hardware might be better.
  • Ultrasonic humidification: For the correct performance of servers, specific humidity is needed. In comparison to other methods, this technology is able to create the necessary moisture with only a tenth of the energy.
  • Hot and cold aisle containment:

  • Use of natural resources:
    • Water cooling: Using water for cooling, be it by directly cooling servers or air.   This technique is impossible to use without a considerable source of water.
    • Free air cooling: Instead of artificially cooling air, use outdoor air. The success of this technique depends on the natural air temperature (location).
  • Green software: When operating systems and applications are developed, they do not often take into account energy consumption. Although it is true that hardware is the area where the most efficiency can be achieved, the software is the way to go for easy, low cost and mass distributable solutions: the creation of green compilers and green operating systems are examples of that.
  • Monitoring power consumption: If an administrator of the datacenter has information about how the resources are used, he can try to reduce its usage implementing energy-saving measures. Monitoring, in addition, can be used to automatize this process, making the energy consumption “dynamic” depending on the actual necessity. One such example is Google, which uses machine learning to save energy.

That being said, focusing on green datacenters may be perhaps the wrong focus: the root of it, after all, is the generation of non-sustainable energy. As such, it might be simpler and more logical to link green datacenters, as it is often done, to the use of energy-efficient technology.

In other words, it does not really have to be about “green solutions”. As we have said, green datacenters are intrinsically linked to a decrease in the consumption of energy, which means, in the context of datacenters, that pressing for green solution automatically lowers energy costs. Taking into account that energy consumption takes a great chunk out of a datacenter monthly budget, more often than not, implementing green techniques will be worth it regardless of the implementation cost of the green technique.


Add new comment

This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.
1 + 7 =
Solve this simple math problem and enter the result. E.g. for 1+3, enter 4.