What do Climate Change, Data Centres and Cloud have to do with each other?

Paul Johnston
4 min readOct 2, 2018

Last week I did a talk titled Serverless is the Future (or is it?) (link to video of talk) and as part of the talk, I launched a whitepaper that Anne Currie and I co-authored. That whitepaper is here:


We have produced this whitepaper after becoming aware of the huge energy resources that Data Centres and the Cloud use to power their services, and that’s without mentioning the huge energy used by cryptocurrency mining which is frightening. These energy sources produce carbon emissions that produce Climate Change.

Currently Data Centres and Cloud are at least 2% of global emissions

2% of global emissions just from Data Centres

So the tech industry produces at least 1/50th of all carbon emissions and that’s likely to grow by up to 5x in the next 7 years.

That’s a bigger problem than the aviation industry.

Tech, “efficiency” and the apathy of the many

When we started to research these issues, one of the things that we found was that the tech world was almost completely unaware of this.

The story around sustainability for Data Centres and cloud providers is often one of “efficiency” — making Data Centres use their energy in better ways, or utilising their energy in better ways.

But here’s the thing.

If you make something more efficient, then the rate of consumption of that resource increases (if you spend any time reading/listening to Simon Wardley you’ll know this is called “Jevons Paradox”).

So by making their Data Centres more “sustainable”, the vendors and providers are increasing their energy requirements by increasing usage.

In one sense, making better use of the energy is a good thing.

But there is another sense, that the revolution in cloud — the ability to startup lots of different types of compute on demand — has led to an explosion in usage, and a massive increase in demand on the energy grid.

But it doesn’t feel like we’re making those energy demands, because running a few API calls from a command line seems like it is basically emissions free.

Except it isn’t emissions free.

API calls to cloud services have emissions attached.

That explosion in cloud? It needs energy.

More energy needed = more energy generation required.

And we simply haven’t built renewable technologies to cope with the demand.

So that energy has come ever more from unsustainable sources, even though some companies are offsetting up to 100% of their energy usage.

And we haven’t built Data Centres to properly consider Climate Change.

And on top of all of this, almost no developers and technologists are even aware of the impact that tech is currently having, and likely to have over the next few years.

It feels very much like the lack of awareness is leading to apathy — “It’s not my problem”. Well, it is our problem. It’s all our problems.

Why Anne and I want Sustainable Servers by 2024

We are both technologists, and want to raise awareness of the issue of Data Centres and Climate Change among other technologists so that we can do something about it.

Believe it or not, I started to use Lambda in 2015 because I wanted to use less compute resource because I care about the environment. I also used the AWS region in Ireland because it is sustainably offset (I reckon many of you won’t even know that! If you did not, read the whitepaper).

So as a first step Anne and I have created a pledge.

It is a simple pledge.

We want all Cloud and Data Centre providers to use Sustainable energy for their Data Centres by 2024.

Signing a pledge is a simple step and we think a “no brainer”. Why would anyone not sign it?

We will take your pledge to the cloud providers and talk it through with them.

The tech world has spent a long time thinking of itself as highly innovative. It is. But it is not facing it’s responsibilities.

Climate Change is the biggest issue we face, and some of the biggest companies in the world are tech companies, and they are becoming part of the problem.

We can fix this. But we can’t fix this without coming together, talking about the issues and innovating.

Oh… one more thing

There are no easy answers here.

And don’t put it off thinking that “somebody will crack fusion power” (or something else like it) because I heard that said 20 years ago, and nobody has done it yet. I reckon it won’t be done in another 20 years either.

Sometimes we have to make hard choices and those choices affect the way we do business.

This is a big problem. It is a hard problem. It is not a problem that “other people” will solve for us.

We can do this. We’re making it as simple as possible to get involved. You don’t even have to show up — just make your feelings known and we’ll do the rest.

This is a hill worth dying on



Paul Johnston

ServerlessDays CoFounder (Jeff), ex AWS Serverless Snr DA, experienced CTO/Interim, Startups, Entrepreneur, Techie, Geek and Christian