Podcast: Sustainable data storage and how to achieve it

What is driving the need for sustainable storage?

At the high level, it’s really about people, profit or planet, because of sustainability. Customers mostly have a green agenda, and they’re trying to deliver on that. And the planet needs to have more sustainability in general.

But why storage in particular? There have been a couple of surveys. Seagate did one that analysed data stored across various storage devices, and they saw that 68% of that data isn’t used at all once it’s created.

So, we’re storing a whole load of data that we don’t have to, and the consequence of that from a storage perspective is we’re generating power and cooling requirements, and we’re also paying for that storage.

So, there’s a lot of focus on, if we are going to store the data, how do we make that more sustainable through the storage platforms we deploy?

How are storage array makers making storage more energy efficient?

There are various things. When you look at a storage array, it’s not just about the hardware. It’s about the hardware that provides the data storage, but it’s also about the software that sits around it. And then it’s about what you do with the data you put on there.

So, at the base level, from a hardware perspective, storage arrays nowadays, we ship titanium power supplies with our storage. So, you’re getting the most efficient power supply units for hosting that storage. And that’s just about delivering the best hardware, the most sustainable hardware you can.

But you still have to store the data. And the next wave we’ve seen is around the use of things like capacity [QLC] flash. That type of technology makes a massive difference in terms of the sustainability of the storage array because you can get much better density, often 10:1 density compared to spinning disk arrays, for example.

That density means you’re able to store more data, but consequently, you have less power requirements, less cooling requirements. You take up less floor space as well. So, the whole kind of environment that you can deliver from a storage array perspective can make it much more efficient from a hardware perspective.

When you look at the software the storage arrays provide to manage that data, the first thing they can deliver is storage efficiencies. It’s not unusual nowadays to see an average of 4:1 to 5:1 efficiencies in terms of the amount of physical tin to the data you’re storing. Those efficiencies drive less power and cooling because you need less tin to host the data.

But then, when you actually host the data, there are other options you can bring in from the storage array perspective to make that even more efficient.

For example, if you look at where the data is hosted, maybe the cloud is a greener option because of their better PUE [power usage effectiveness] datacentre ratings. If you can leverage data tiering to tier from the storage array and leverage the cloud as part of the overall data mesh or strategy, then you can leverage the greener cloud capabilities to store your data.

The final piece is if the storage array itself can bring data optimisation technology. So not just storing the data more efficiently, but actually analysing what the data is and then giving the customer guidance on how they could reduce the amount of data they’re storing – for example, duplicate data, stale data, data that’s not been accessed for a while.

If your storage array can give you that kind of insight, you can actually start to delete data, which is something we’ve never done before, and achieve optimisations that way as well, ultimately all driving better energy efficiency.

What benefits to IT and the business do more energy-efficient storage arrays bring?

I think the biggest one to the business is probably cost, because if you’re more efficient in how you store data you need less infrastructure to do that. So that reduces the cost the business has to pay, and obviously IT has to fund from that.

The consequence of that cost and how you look at it is that you’re potentially paying for less hardware to host the same amount of data, but you’re also significantly reducing your energy bills.

As we saw during the last energy crisis, that has a big impact on what businesses are having to pay to store their data. It is estimated that by about 2030, 38% of the datacentre energy bill will be down to data storage.

So, it has a huge impact on the cost, but then also from a business perspective being more optimal in what you do improves the floor space. Maybe you can reduce datacentre footprint. Maybe you can reduce the number of datacentres.

And then, of course, that improves the planet’s posture. Why power and cool all of this data storage if we don’t need to?

That has a knock-on effect to everybody else, not just to the business itself. There are a lot of advantages from putting in more efficient storage arrays at the very bottom of what’s driving data hosting and the energy generation to support that.