Originally posted on Data Center POST

Byline: Steve Friedberg

The Data Center World conference always has its share of optimists; experts who see significant improvements in the industry coming over the near term. But they’re also realists. They understand that not everything happens as quickly as they’d like, if it happens at all.

In that context, Data Center POST asked a group of vendors and industry analysts to play the role of “grump,” predicting what could happen in the next several years that might slow down the pace of implemented innovation.

For instance, Joris te Lintelo, SVP of maincubes secure datacenters, sees power challenges as a gating factor. “I think Western Europe is slightly different than the US where you have your own power constraints. But I think in Europe, it’s a mess at the moment. There’s no proper planning. We are really out of capacity and power,” he told Data Center POST. “Politicians don’t have to put a stop on development because there’s simply no availability of power. That is a big challenge in the Netherlands, but also in the FLAP-D markets. It depends on how you look at geographies; there is an enormous amount of power not too far from Dublin. We see the same in Germany, where we have the opportunity to build up outside of the cities outside of Berlin and Frankfurt, but it’s a challenge within the metropolitan areas.”

Andy Abbas, the CEO of TFORM, Inc., has identified potential challenges due to inaccuracies in IT operational data that could lead to significant issues for companies looking to move their workloads back from cloud environments to on-premises settings. “Clear visibility into your IT asset management and their dependencies is crucial for organizations contemplating repatriating their data and applications to on-premises infrastructure,” he explains. According to Abbas, there are several reasons why companies might consider such a move, including concerns over cost, security, or performance. These factors are often the primary drivers behind the decision to transition workloads back to an on-premises environment, he notes. “Companies facing inaccuracies in their IT asset management should take the time to resolve these issues before starting their efforts to move workloads back from the cloud,” he told Data Center POST.

Kirk Killian, the president of Partners National Mission Critical Facilities, likewise urges a more methodical approach, saying companies that rush toward adoption face a bevy of potential challenges. “Not unlike anything else in life, there are planners and then there are non-planners. And we see the companies that do a very good job of the scoping, the planning the methodical approach to where should we put these applications and data are probably going to come out ahead,” he states. “The other one is that instead of ‘ready, aim, fire,’ they take a ‘ready, fire, aim’ approach; we will often find you’re going to not get the result that you would have wanted to get.” Killian also says that in some cases, “companies will hire a new 35-year-old whiz-bang CTO from Los Gatos, California, who says in almost a knee-jerk reaction, ‘we’re doing it this way,’ before really studying what is the best way to do it. For a lot of those, we see that there are some hiccups along the way.”  Killian stresses, however, that he believes there’s a prominent role for public cloud implementation.

Artificial Intelligence (AI), of course, was a huge topic at Data Center World, as it is virtually everywhere these days. 451 Research’s Perkins Liu is urging companies not to rush headlong into adoption. “Everybody is all AI. People are still investing into it. A lot of people are optimistic about it, but we still need to see whether it will eventually be beneficial to our life, our industrial productions and all aspects of human life and social life and economics. Somebody has to pay the bill eventually. I think we’re getting there. But like the cloud, it has taken years for us to be realized, right? We have seen 5G emerging, but it didn’t really happen in metaverse. It was exciting at the beginning, but it didn’t really happen to what we expected. Right now, AI seems to be on the track. But the only challenge I see is, do we have enough power to support that kind of load? If at the end of the day, people are saying at the end of AI is power, is energy; if we don’t have enough energy or power to support, there’s nothing we can do.”

To continue reading the full article, please click here.