The public cloud is the greenest cloud. The three hyperscale providers, Amazon Web Services (AWS), Google and Microsoft outperform the other types of data center providers in energy efficiency. This gap is large and it is not expressed as a percentage, but rather as a multiple. Furthermore, this gap will increase. As a consequence, companies can reduce their carbon footprint in IT most by migrating to a public cloud service provider and for many companies this reduction will be by an order of magnitude. For example, AWS offers a broad range of components and services. The company also provides detailed descriptions how these are deployed to maximise the reduction of your carbon emissions.
Energy consumption has flattened despite sharp increase in use of IT
Energy consumption and carbon emissions related to IT have been a topic of concern this century. The ubiquity and the growing importance of IT in our professional and private lives resulted in gloomy projections. A study sponsored by the US Department of Energy in 2016 traced the data center electricity consumption back to 2000. In the first decade of this millennium the increase was significant: 90% from 2000-2005 and a further 24% from 2005-2010.
However, the study found a “drastic increase in demand for data center services” combined with a “near constant electricity demand” since 2007. The report explains that this flattening “coincides with the rise in very large “hyperscale” data centers. Driven by an increased popularity of moving previously localized data center activity to colocation or cloud facilities.”
Moreover, the report concluded that energy consumption would even decline if the shift “towards more server use in large data centers” would accelerate. In such a “hyperscale shift scenario” an estimated 45% reduction in electricity demand would be feasible.
Hyperscale providers energy reduction outweighs further growth in IT
A paper published in Science illustrates the impact of our ‘digital lifestyles’ in the second decade of this century:
“Since 2010 […] the data center landscape has changed dramatically. By 2018, global data center workloads and compute instances had increased 6.5-fold, whereas data center IP traffic had increased 11-fold. Data center storage capacity has also grown rapidly, increasing by an estimated factor of 26 over the same period.”
This study also quantifies the development of data center energy use. In particular, the hyperscale data centers fueled a spectacular improvement of the average electricity efficiency:
“Since 2010, electricity use per computation of a typical volume server […] has dropped by a factor of 4. At the same time, the watts per terabyte of installed storage has dropped by an estimated factor of 9. Furthermore, growth in the number of servers has slowed considerably owing to a 5-fold increase in the average number of compute instances hosted per server, alongside steady reductions in data center power usage effectiveness. These trends have been largely driven by shifts in compute instances to energy efficient cloud and “hyperscale” data centers.”
Looking ahead, the study predicts how much longer these efficiency trends can last. This is a relevant question because “the next doubling of global data center compute instances may occur in the next 3 to 4 years.” According to these experts, we can expect further efficiency gains if the shift continues. Moreover, their conclusion is positive: “there is a sufficient energy efficiency resource to absorb the next doubling of data center compute instances.”
Continuous innovations from a CSP like AWS deliver tangible reductions
For a cloud service provider (CSP) like AWS it makes sense to reduce the energy consumption. In fact, every IT service or software provider who’s responsible for delivering ‘as a service’ benefits from a reduced energy bill, but for a CSP like AWS who offers their services on a pay-per-use basis a lower consumption enhances the capability to compete with a lower price.
As a consequence, AWS cares about developing more energy efficient components and services. Increasingly AWS also shares best practices how to maximise the impact for your situation. For example, in 2021 AWS published detailed descriptions on how to optimise your infrastructure specifically for sustainability in the areas of compute, storage, and networking. A more recent blogpost contains a list of possible measures that can be put in place. Measures that range from selecting more energy efficient hardware (by using processors such as AWS’ own Graviton) to intelligent deployment optimisation services (by using services such as AWS Compute Optimizer).
Migrating to public cloud will reduce your carbon footprint significantly
Several independent researchers also looked at the carbon footprints of the individual companies. Comparing the performances of Amazon Web Services, Microsoft and Google with traditional data centers. These companies also publish frequently about their innovative approaches to become ever more ‘green’. Many examples are given. About deploying AI to reduce the cooling bill autonomously. How switching to hyperscale SaaS offerings is much more efficient. Publications describe investments to replace diesel powered power failure equipment with solar powered battery installations or with hydrogen fuel cells. Their efforts extend beyond electricity use and include investments to reduce water consumption.
The results are impressive, even when we set aside the impact of their renewable energy purchases at scale and their direct investments in renewable energy across the world. Every one of these companies is on its way to be carbon neutral ahead of the timeline set in the Paris Climate Agreement. Some aim already beyond neutrality and are planning to become carbon ‘free’.
In conclusion, companies can reduce their carbon footprint in IT most by migrating to public cloud. Moreover, as the data center industry continues to improve its energy efficiency this is another area where these providers will increase their lead.
When you want to discover how you can reduce your carbon footprint significantly and quickly? Get in touch and send us a message.