Circular economy of the cloud: The whole is greater than the sum of its server parts

AWS Reverse Logistics FA Testing Lab80 1024x576 1


Author: Ken Haig, Head, Energy and Environmental Policy, APAC, Amazon Web Services

As we continue to innovate and empower Australia towards a digital future, it is important that we are investing and innovating in sustainability to help create a more sustainable future.

Everything that happens online today is facilitated by data centres. Incredibly, innovation and investments in energy efficiency and renewable energy by cloud service providers in particular continues to lower the carbon footprint of much of this computing in the cloud. In fact, 451 Research found that migrating IT workloads from on-premises enterprise and public sector data centers to cloud data centers in the Asia-Pacific (APAC) region can reduce energy consumption and associated carbon emissions by nearly 80%.

In addition to energy efficiency, Deloitte has found that the sustainability maturity of data centers should be assessed against renewable energy, water usage, and infrastructure circularity. Although infrastructure circularity is not as widely understood, it offers meaningful opportunities to simultaneously decrease emissions and cost through continuous improvement and waste minimisation.



Circular economy thinking and principles

Giving retired data center hardware a second chance and extending hardware life also supports net-zero carbon goals. At Amazon Web Services (AWS), we embrace three circular economy principles for our server racks that may be helpful to consider:

  1. Design – Server rack circularity and sustainability must start with design. This means focusing on eliminating excess and unnecessary materials such as steel or plastic, and increasing use of recycled and biobased content. It also means prioritising designs for repair, reuse, and recycling right from the start to ensure the feasibility of these processes throughout a server rack’s life cycle.
  2. Operate – Using equipment for as long as operationally efficient reduces the carbon footprint associated with hardware use. Extending equipment lifetime prevents materials from becoming waste, which helps reduce the amount of e-waste that is produced. This is why, in 2022, AWS extended the life of our servers from four years to five, and that of its networking equipment from five years to six.

Complementary to this principle is ensuring a robust maintenance and repair program is in place. This helps to increase component reuse, and further reduce carbon emissions and waste across the supply chain. For example, regular maintenance can help to extend the lifespan of components such as production racks and networking equipment.

     3. Recover – When it’s ultimately time for server racks to be decommissioned, sanitised equipment can still be routed for further component reuse. This involves testing, repairing, and recirculating non-media storage equipment back to data centers, or even for reuse by third parties.

Throughout the process, security must remain at the core of the program. Cloud service providers, such as AWS, have controls and processes designed to securely and carefully manage media storage devices used to store customer data. This includes procedures on how to install, service, and eventually sanitise the devices when they are no longer useful. Media that stores customer data should not be removed from control until it has been sanitised.

Only when this sanitisation process is fully completed can any hardware then be recycled to recover raw materials, which in turn can be reincorporated into new products.

The final stage in a device’s lifecycle: Recovery and reuse

As the last stage in a device’s lifecycle, component reuse is important in ensuring minimal waste generation, and reducing the carbon emissions, and extraction of critical raw materials.

After data storing media is sanitised and all data has been securely removed, decommissioned server racks should be robustly tested for functionality. This means exhausting feasible avenues in testing and repairing used components, while maintaining care and security throughout the process.

The next step is the de-manufacturing process, where server racks are disassembled into individual components, with each component processed and evaluated for its highest reuse potential. Components prioritised for reuse can range from power supply units (PSUs) to graphics processing units (GPUs), and fibre optics. These individual components are then sent for repairs and testing, including failure analysis testing and testing rigs, to ensure they function as intended. The data from these tests are also useful in identifying hardware or software modifications, which can then be fed back into the design stage to improve new product development and further extend component lifespans.

The final step in all this is the processing and reuse of functional components – parts that can go back to the data centers they came from, to continue powering cloud computing that drives our world today.

The whole is greater than the sum of its server parts

Understandably, this may seem like a lot of thought, consideration, and work for a single component in a server rack, in one data center out of the fleets of data centers that cloud service providers manage. But it is ultimately important that businesses are actively committed to, investing in, and innovating for sustainability to create a more sustainable digital future. We do all of this in our AWS Reverse Logistics hubs around the world and continuously innovate to meet sustainability goals for our customers.

The whole, in this case, is greater than the sum of its (server) parts. Every single component that is tested, repaired, cleaned, and reused may be small, but if the useful life within each of these can be extended, it will culminate in a tremendous impact for people, businesses, the cloud computing industry, and ultimately our planet.

  • Oceania Luxury Travel Co Luxury Travel Australia Banner 728x90 1