The heata team were at Civo’s Navigate in London this week, with our Co-Founder, Mike, speaking on the Sustainability Panel. There were lots of interesting discussions, but here are Mike’s top 3 takeaways: 

Scope 3 emissions are the elephant in the room

The conversation often focuses on the emissions related to the power consumption in use. But what about all the emissions associated with the data centre's construction? This includes both the building/infrastructure and the components within.

This is an argument for thinking more critically about where a new data centre’s infrastructure should be located. One criteria could be existing buildings where there is a use for waste heat. e.g. pre-existing heat networks, businesses that have a specific need for heat. Or within buildings to support core heating needs. At heata we take that principle a step further into people's homes to supply free hot water, and we also use second life CPUs which extends their lifecycle - another pillar of Scope 3 reporting.

It was great to see one of the Open Compute Project's rack standards at Vespertec's stand. It got us thinking that something similar which builds a framework around physical infrastructure (e.g. reuse first rather than new build) would be helpful too. Perhaps this exists, but if it does it’d be good to hear more about it.

‍

We need carrots, sticks and greater transparency to encourage change

Some quite big regulatory sticks (CSRD and EED) are going to force both data centres and companies above a certain size to report their carbon impact in greater detail, going beyond emissions into broader environmental and social effects. It'll be interesting to see what sort of a scramble this initiates, still there's not so much in the way of carrots to help companies make a shift away from comfy incumbency. What sort of incentives could help there?

Greater transparency and a good framework for understanding and reporting the impact of cloud compute will also really help reduce greenwashing and enable comparison of different options on a level playing field. It was great to meet greenpixie who are helping to do just this. (Read more about them here >>> ).

We discussed one of the fallacies of what is currently called 'sustainable' compute. Some companies report Scope 2 (e.g. the emissions that are associated with the power used by a data centre) using a market-based approach. Through contracts (e.g. renewables and offsetting) they are able to report no emissions (100% renewable!), which is VERY misleading. Location-based reporting shows the actual carbon intensity based on the location (geographic locations can vary significantly, roughly speaking; London: bad, the North: better), which can then be negated..but at least there is visibility on the reality.

‍

What data centres enable is amazing (and they could do good beyond their core purpose)

There can be a tendency to cast data centres as the bad guys but it's good to be reminded of how fundamental they are to our modern lives. We had some inspiring conversations at Civo Navigate which was a real reminder of the innovation and progress that they support, and we came away thinking that there really is a big opportunity for data centres to have a positive impact beyond their core purpose.

Given the scale of the investment in compute infrastructure to come (particularly with the explosion in AI - we had a great chat with Jason Grant at Nvidia) the opportunity now is to rethink the how?! Rather than stopping housing being built, and hoovering up water supplies, they could be supporting local communities by providing free hot water to households, swimming pools and local businesses. And even better it could be compute from local businesses supporting their own community. Or even AI requests answered by local compute networks! Sometimes it's good to imagine what we want the future to look like and then work back from there.