With so much focus globally on corporate sustainability commitments, it’s unusual to see carbon emissions being a critical factor of influence in IT procurement. Speaking to my colleagues in the MSP sector, few RFPs are focused on reducing carbon emissions when evaluating moving to the cloud against remaining in their enterprise data center.
That being said, many enterprises do have sustainability targets. A significant contributor to an organization’s carbon footprint is the energy consumption related to the data centers hosting their enterprise applications. For organizations still supporting their own enterprise data center, the likelihood is that these facilities have not been built, nor operate, with carbon efficiency in mind.
The proof
Multiple studies have shown that moving from enterprise data centers to the cloud can reduce carbon emissions by over 95 percent. For example, a 2018 study commissioned by Azure in partnership with WSP concluded that Azure can be up to 98 percent more carbon efficient than traditional enterprise data centers. Similarly, studies performed by 451 Research found that moving on-premise workloads to AWS can reduce carbon footprint by up to 96 percent.
The ”Big 3” hyperscalers, AWS, Azure, and GCP, each have their own published sustainability initiatives. For example, AWS has committed to net-zero carbon emissions by 2040. Azure’s commitment is to be carbon-negative by 2030 whilst GCP has committed to running on carbon-free energy 24/7, such as wind and solar, by 2030.
They achieve these reductions primarily through better resource utilization, optimization of data center design, efficient cooling systems, power-efficient chipsets, and procurement or generation of renewable energy. These opportunities are just not available to traditional enterprise data centers.
Cloudy with a chance of carbon savings: Regional footprint insights
From a customer’s perspective, the added benefit of moving to the cloud is the ability to gain real-time visibility of their carbon footprint through the carbon emissions reporting they each provide. In addition, when planning migrations to the cloud, businesses can get a predicted emissions calculation based on their specific design and preferred Region. Even within each hyperscaler, each Region has a different level of sustainability. When comparing the carbon emissions for equivalent AWS workloads in South Africa vs Dublin vs Sweden we get a lower carbon footprint solution when comparing hosting in Sweden to Dublin, and materially better when compared to South Africa.
Conclusion: Reduce the environmental impact and save more with the cloud
Making the correct hosting decision is merely the first step toward reducing carbon emissions. Once operational on the cloud, it is crucial to consider carbon optimization of the entire estate. Right-sizing to smaller instances, using stop/start schedules, and archiving and compressing data not only lower operational costs but also reduce the carbon footprint.
Given the significant opportunities to positively impact a company’s sustainability targets, a case is made for businesses to align their IT procurement and decision-making processes with their corporate sustainability commitments. This includes incorporating estimated carbon emissions and optimization processes as part of their evaluation criteria.
