Schneider Electric CTO claims AI and cloud could give DCIM market new lease of life

Embracing cloud, big data analytics and artificial intelligence (AI) could help suppliers breathe new life into the under-performing datacentre infrastructure management (DCIM) market, claims Schneider Electric chief technology officer (CTO) Kevin Brown.

In a keynote presentation at the Datacloud Europe Congress in Monaco, Brown told delegates about the challenges the DCIM market has faced over the past decade or so, as adoption of the technology has failed to keep pace with the hype generated about it by analysts and suppliers.

Examples of this hype include multibillion-dollar projections about the amount of money DCIM software could save the datacentre industry in efficiency gains and the like, but these predictions have never quite come true.

“We had all this promise and hype about what these tools were going to do, and this led to some fairly high industry expectations,” said Brown.

This, in turn, led to inflated expectations and predictions from the analyst community about how big a market it could grow into, with Brown citing figures that, at one time or another, had its valued pegged at between $3bn and $7bn.

“It’s roughly now [worth] around $700m,” he said. “We were expecting [market value] anywhere between $3bn and $7bn, and it actually came in now at $700m.”

This is despite the fact that many of the issues DCIM products were created to address still exist within datacentres today. The problem is, said Brown, the products were not built to address the issues in a way that appeals to users.

For example, DCIM tools can often be difficult to use, hard to maintain and can leave users at high risk of information overload and alert fatigue, as they are bombarded with notifications about irrelevant metrics, he said.

Part of the reason for that is that suppliers have sometimes struggled to understand who the target audience for DCIM tools are – not just from an enterprise, colocation or hyperscale datacentre perspective, but also from an IT department or facilities management point of view.

Following on from this, some of these products ended up having functionality or tracking metrics that were only of use to a relatively small subset of users, which has also served to limit their appeal.

As far as Schneider Electric is concerned, there is a still a clear need and demand for DCIM tools, particularly those that are as intuitive and easy to use as setting up an iPhone, said Brown.

Operators still need to keep tabs on how their infrastructure is performing from an uptime, availability and energy consumption perspective, which DCIM can help with, he added.

“The problems still exist, but our conclusion is that the tools did not match or solve the problems people had,” said Brown, but adding cloud, big data analytics and AI tools into the DCIM mix could change that, he pointed out.

“We have things happening with analytics and AI, and we’re still doing the manual mapping of [datacentre] assets,” he said. “It’s a very complex problem for most customers, and we have teams dedicated to solving that with AI. But why is it we still require people to be manually involved with where devices are in the datacentre?”

There is also potential for AI to change the role that DCIM plays, so that it is more proactive in datacentre management, with Brown floating the idea of it being used to enable the predictive maintenance of datacentre assets. After all, DCIM has traditionally had a reputation for being more of a reactive technology.

“Ultimately, our view is that traditional DCIM tools are solving these two problems: what happened and why did it happen?” he said. “And that is simply not good enough to justify a $7bn market and this next generation of tools that we see coming are going to be about what will happen and how can I prevent it.

“We need to address all four of these problems if we want to realise the promise and potential that DCIM has.”