The next significant revolution — Digital Decarbonization

In the last ten years, companies have placed their attention on digital transformation to enhance productivity, connect with customers promptly, and use the most advanced technology to offer goods and services. The integration of cloud technology has played a crucial role in digital transformation by enabling businesses to transfer their workloads to the cloud, gaining benefits such as flexibility, scalability, and cost-effectiveness.

However, the growth of digital transformation has also come with an environmental cost. The increased demand for compute, storage, and processing power has led to a surge in carbon emissions from data centers. This is a major concern for businesses and governments around the world, as climate change is a growing threat.

There are a number of ways that businesses can reduce the carbon footprint of their digital transformation initiatives. One way is to adopt sustainable cloud computing practices. This includes using cloud providers that use renewable energy sources and implementing energy efficiency measures in cloud deployments.

While those are excellent ways to get carbon emissions under control, further improvements can be made across the enterprise level to standardize how digital products are developed and delivered to customers, which can also help to control carbon emissions.

Changes and improvements across enterprises can be made in 5 major areas -

1. People
2. Process
3. Data
4. Infra
5. Energy

Pillars of Digital Decarbonization

Let us see how we can improve in these areas to help save the planet.

1. People — They are at the heart of any organization, and their awareness and commitment to sustainability play a pivotal role in reducing carbon emissions. Employees, stakeholders, and consumers must be educated about the environmental impact of their actions and encouraged to adopt eco-friendly practices. This can be achieved through workshops, training sessions, and awareness campaigns highlighting sustainable behaviors’ importance.

Furthermore, fostering a culture of innovation and sustainability is key. Enterprises can empower their workforce to come up with creative solutions that minimize carbon footprints. Encouraging cross-functional collaboration can lead to the development of groundbreaking ideas that positively impact the environment.

As of today, there are no direct guidelines or tools that tell us what impact digital products have on the environment. By adopting big cloud providers like Google, for example, we can at least be sure that choosing a low CO2 region and running the workload from there will result in low CO2 emissions.

Besides, developers should know more about the ‘Unit cost’ of running a code, module, or library. The unit cost helps the developers understand the importance of writing effective code.

Carbon pricing is an effective method of informing a team about their carbon footprint. In the future, cloud providers could potentially disclose the CO2 emissions associated with the team’s digital products.

2. Process — It is a set of interrelated tasks to achieve a specific outcome. In the context of people and enterprise, a process can be defined as a series of steps to create, deliver, and improve products or services.

To me, 2 areas stand out concerning building a quality process to bring in a great culture with sustainable development practices

FinOps — FinOps can help developers limit carbon emissions in several ways, including:

• Providing visibility into cloud costs: FinOps can help developers understand how their cloud resources are used and where costs are incurred. This information can be used to identify opportunities to optimize cloud usage and reduce costs.
• Encouraging sustainable cloud computing practices: FinOps can educate developers about sustainable cloud computing practices, such as optimizing resource utilization and automating cloud deployments.
• Collaborating with other stakeholders: FinOps can work with other stakeholders, such as procurement, IT operations, and security, to develop and implement sustainable cloud computing policies and procedures.

Learning at the source — Federated learning and on-device learning can help in limiting carbon emissions in some of the following ways:

• Reduced data movement: Federated learning and on-device learning do not require data to be moved to a central server for training. This can save energy, as the data does not need to be transferred over the internet or stored on a centralized server.
• Reduced computational resources: Federated learning and on-device learning can be performed on devices with limited computational resources, such as smartphones and laptops. This can save energy, as the devices do not need to use as much power to train the models.
• Reduced cooling requirements: Federated learning and on-device learning do not require the use of powerful servers, which can require a lot of cooling. This can save energy, as the servers do not need to use as much power to cool down.

3. Data — Businesses must understand customer patterns, interests, and behaviors. However, the amount of data generated and stored in big data centers is staggering. Some of this data is single-use, meaning it is only used once or not at all.

The outputs from on-device learning, such as formulas, patterns, and machine learning models, can be stored in cloud data centers instead of storing large volumes of raw data. This eliminates the need for a lot of energy to power the storage modules.

Eliminating single-use data can significantly reduce carbon emissions from the tech industry. In fact, I sometimes think of single-use data as the equivalent of single-use plastics. Just like we need standards and laws to eliminate single-use plastics, we also need standards, guidelines, and best practices to eradicate single-use data.

Here are some specific ways to eliminate single-use data:

• Use federated learning or on-device learning to train machine learning models. These techniques do not require data to be moved to a central server, which can save energy.
• Use data compression techniques to reduce the size of data sets. This can also save energy, as less data needs to be stored and processed.
• Delete data that is no longer needed. This can be done automatically by setting retention policies for data sets.
• Use data anonymization techniques to protect privacy. This can help to reduce the amount of data that needs to be collected and stored.

4. Infrastructure — When it comes to digitalization, cloud providers offer top-notch infrastructures. However, this comes at an environmental cost. To reduce energy consumption at cloud or data center regions, let’s explore some solutions.

• Avoid storing single-use data: Single-use data is only used once or not at all. Keeping this type of data can waste energy and resources. Enterprises can avoid storing single-use data by using data compression techniques, deleting data that is no longer needed, following data retention policies, rethinking how data has to be stored, and so on.
• Determine if a solution truly requires high availability across multiple regions and zones: High availability is the ability of a system to remain operational even when there are failures. However, high availability can also come with a high environmental cost. Enterprises should carefully consider whether a solution truly requires high availability across multiple regions and zones before implementing it.
• Utilize cold storage and hot storage appropriately: Cold storage is a type of data storage that is used for data that is not frequently accessed. Hot storage is a type of data storage that is used for data that is frequently accessed. Enterprises can save energy by storing data that is not frequently accessed in cold storage and storing data that is frequently accessed in hot storage.

5. Energy — More than 50% of world energy comes from coal and gas, which contribute to global CO2 emission. All the major cloud providers across the globe have committed to consuming clean energy for their data centers. Huge investments have already been made in this area. We should continue to invest in clean energy and demand that all cloud providers use clean energy.

• Support policies that promote clean energy: We can support policies that promote clean energy, such as tax credits for renewable energy and renewable portfolio standards.
• Invest in clean energy technologies: We can invest in clean energy technologies like solar, wind, and Hydro power.
• Demand that cloud providers use clean energy: We can demand that they use clean energy by contacting them directly and supporting organizations working to promote clean energy in the cloud.

Please share your ideas on how to reduce carbon emissions.

Happy learning!

Originally published at http://shankarkumarasamy.blog on August 28, 2023.