How has the recent turmoil within the OpenAI offices changed your plans to use GPT in a business process or product in 2024?
Increased uncertainty means we are more likely to evaluate alternative AI chatbots and LLMs.
No change in plans, though we will keep an eye on the situation.
With Sam Altman back in charge, we are more likely to go all-in with GPT and LLMs.
What recent turmoil?
Cloud Native Ecosystem / Cloud Services / Kubernetes

Can Reducing Cloud Waste Help Save the Planet?

By reducing over-provisioning and cloud waste in a green cloud optimization effort, you can cut your carbon footprint and cloud bills.
Sep 24th, 2021 10:00am by
Featued image for: Can Reducing Cloud Waste Help Save the Planet?
Featured image via Pixabay

Annie Talvasto
Annie is an international technology speaker, CNCF Ambassador, Azure MVP and specialist in the Kubernetes and cloud native ecosystem. Annie hosts and produces the Cloud Gossip podcast and has been a co-organizer of the Kubernetes and CNCF Finland meetup since 2017. She works at the Kubernetes cost optimization startup, CAST AI.

Enterprise spending on cloud infrastructure is growing faster than ever. In 2020, it reached almost $130 billion, a 35% jump compared to the previous year.

When looking at it from the perspective of sustainability, the move toward the cloud seems to be a good idea. After all, cloud services are usually more efficient than traditional data centers. The UK government is officially advising companies to move to the cloud to curb carbon emissions as part of the country’s broader campaign for cutting carbon footprints in half by 2030.

Few companies, however, consider the sustainability implications of their cloud setups with their choice of cloud provider. As a result, they’re missing out on a huge business opportunity. By eliminating over-provisioning and cloud waste in a “green cloud optimization” effort, businesses stand to lower both their carbon footprint and their cloud bills at the same time.

The First Step is Understanding the Context

Why do we need to make cloud computing greener? Let’s take a look at the energy requirements and carbon footprint of the information and communications (ICT) technology industry.

The amount of energy data centers require doubles every four years. Every new region opening from AWS, Azure, or Google Cloud Platform contributes to this. According to Nature, in one hopeful scenario, the ICT industry will be responsible for 8% of total electricity demand by 2030, an increase of 15 times from 2010.

What about the global greenhouse gas emissions generated by ICT? We’re looking at a consistent 2 to 6% of emissions added each year since 2007. Surprisingly, the aviation industry generates just as much.

Modern technologies like artificial intelligence and machine learning require massive amounts of computing power to deliver on their promise. A study of natural language processing models showed that training one NLP model generated carbon emissions similar to 125 round-trip flights from New York to Beijing.

But here’s some good news: Cloud service providers care about energy use, and their hyperscale data centers help make the cloud more energy efficient. Still, to meet the objectives of the Paris Climate Agreement, the ICT sector will have to reduce its carbon emissions by 45% within the next decade. As you can see, there’s still a lot that needs to be done in this area. Paying attention to the cloud’s sustainability should be at the top of enterprise agendas.

Reducing Cloud Waste is a Business Opportunity

The cloud helps teams to scale capacity up and down in line with fluctuating demand, reducing idle time with approaches like serverless computing. By moving to the cloud, the eCommerce giant Etsy cut its energy consumption by an estimated 13%, from 7330 megawatt hours in 2018 to 6376 MWh in 2019 – and all the while, the company continued growing!

Cloud optimization comes with several benefits that affect both your carbon footprint and operational costs:

1. Eliminating Cloud Waste = Lower Cloud Bill

Cloud waste is real and companies of all sizes are struggling to contain it. Shadow IT projects, resources left running or a tendency to over-provision are to blame here.

Right-sizing virtual machines or instances, autoscaling to match workload demands and use of spare computing capabilities (spot instances) are just a few examples of optimization that leads to infrastructure cost and energy savings.

2. Cutting Unnecessary Dependencies Accelerates Time to Market

Making build pipelines more efficient by removing all the things teams don’t need yet still consume resources is a smart move. Remove excess code or instances left running by teams after testing to slash your costs, reduce energy consumption and make developers happier.

This is a perfect use case for automation. Applying all these fixes manually and on a regular basis translates into a significant amount of time, which translates into new expenses. Teams need automation to solve this puzzle since reporting alone is not enough to reduce their cloud bills.

3. Meeting Company Sustainability Goals

Companies that eliminate unnecessary servers and choose cloud providers that use renewable energy cut their carbon footprints and find meeting their sustainability goals easier.

Spotify (take a look at its Sustainability & Social Impact Report) and SiteGround are good examples here. Both switched to the carbon-neutral Google Cloud Platform (GCP) for that reason. Other businesses turn to containers and virtualization to cut the overall power consumption (Forever New did that).

Yet another approach is choosing only those cloud regions covered by renewable energy sources. MapBox does that, in its mission of becoming carbon neutral.

Companies invest in sustainability for many reasons. One of them is attracting and retaining talent, something especially important in the ICT industry plagued by a skills shortage. Even 70% of employees prefer to work at a company with a strong environmental agenda and would stay there over the long term.

Start Your Green Cloud Optimization Project Today

Cloud optimization initiatives not only reduce costs but server usage, as a result cutting energy usage and carbon emissions. Here are a few starting points to get you on the right track:

  • Start your path with visibility: You need to have a clear understanding of the resources you’re using. The market is full of cloud cost-monitoring, -allocation and -management tools to help you get a more granular view of your cloud infrastructure spending. This also helps to make teams aware of the costs and energy consumption they generate and create a sense of ownership about them.
  • Optimize your setup: Use automated cloud-optimization solutions that constantly analyze your infrastructure and automate infrastructure optimizations for cutting costs in line with your policies. They move workloads around for maximum utilization, retire unused instances and choose the right types and sizes of virtual machines to help you avoid over-provisioning.
  • Implement ongoing governance: You need to have processes in place to make sure your teams are dedicated to your sustainability goals related to cloud resource usage.

To achieve the best results, align your cloud optimization initiatives with a continuous software delivery process. This opens the doors to making real progress in implementing green cloud optimization across all of your teams.

Take Action to Reduce the Environmental Impact of the Cloud

You can take action in many ways to combat climate change, from choosing a carbon-neutral cloud provider and selecting data centers in regions that use energy from renewable sources to reducing the number of servers in use and eliminating idle resources.

You can use open source Cloud Native Computing Foundation (CNCF) projects for this purpose. KEDA is a great example here as it scales your event-driven application efficiently. Or, you can optimize your compute usage with the help of autonomous Kubernetes optimization platforms such as CAST AI that search for the most cost-effective and sustainable cloud services.

To find out more, I invite you to attend my session at KubeCon+CloudNativeCon North America 2021: “How Event-Driven Autoscaling in Kubernetes Can Combat Climate Change.”

To learn more about Kubernetes and other cloud native technologies, consider coming to KubeCon+CloudNativeCon North America 2021 on Oct. 11-15.

Group Created with Sketch.
TNS owner Insight Partners is an investor in: Pragma.
THE NEW STACK UPDATE A newsletter digest of the week’s most important stories & analyses.