Colos Not Raging Against the Machine

This statement from Gartner is not fear mongering. It is a timely reminder that not only are colocation providers responsible for facilitating the growth and adoption of the latest technologies - their own survival as businesses is also totally reliant upon them. 

Nor is this a prediction for the distant future - 2020 will fall squarely in the middle of existing business plans. Unless using AI to optimise the performance of data centres is high on the agenda for colos, a challenging, if not bleak future lies on the horizon. 

How to stay in control

As Google has already recognised, in the current climate data centre facilities management has become too complex for humans alone to handle. Artificial Intelligence is going to create vast demand for compute power as organisations look to use Machine Learning and Deep Learning (DL) to gain a competitive advantage in the global economy. This will increase the pressure on colocation providers and their workforce until soon traditional infrastructure and operational models will no longer be capable of enabling such growth.

Google has assessed the number of possible interactions between the different pieces of equipment that constitute a modern data centre and it reached a clear conclusion. The volume of possible configurations would be far beyond the reach of any member of staff to test. If colos are to negate this heavy and complex workload and remain in control of their business, they will require data centre management systems with AI at their core. This will be the only way that they can continue offering unrivalled availability and performance in the face of the pressures and demands imposed by increasingly ambitious customers.

Finding the sweet spot for Artificial Intelligence

The increasing adoption of AI does not mean that we are seeing the dawn of the fully autonomous data centre – yet! What it does mean is that colos are putting themselves in a position to understand exactly how their data centre is functioning. With facilities operators having to manage so much information, whether this concerns data generated by Electrical Power Management Systems all the way through to Building Management Systems, there is a danger that they will be suffocated to the point that decisions stop being made. Smart Artificial Intelligence deployments that are designed to manage data centre facilities and the hardware that drives them are changing this, regularly being used for tasks such as reducing energy usage through to optimising cooling systems. Yet, addressing application within the data centre itself is only half the battle.

Colocation providers need to match these use cases with an integration of AI into wider business strategy. Already there are organisations that are using the technology to form predictive models that allow them to reduce customer churn and foresee challenges amongst their client base. This is the sweet spot that those in the industry have to aim for if they are to maintain and improve their market share. To do this successfully though requires not only technological change, but a shift in the way these organisations work.

Getting ready to integrate AI into the business

Global revenues for AI are due to leap from $2.5bn in 2017 to in excess of $4bn through the course of 2018, according to Statista.  If businesses are to make the most of these investments, they cannot afford to jump into an AI deployment without considering whether their current working structures allow for the successful adoption of the technology. One area that demands specific attention is the role of data analysts and business intelligence professionals. Too often these teams are operating independently of the rest of the business and the organisation either ends up wasting time duplicating effort or missing out on key insights that can be applied for operational benefit. Insight without application will waste resources and, in many cases, limit innovation.

This places colocation providers at a critical crossroads. Choosing the path of integrating Artificial Intelligence into their daily operations will certainly not be a simple process - on the other hand, the consequences of burying heads in the sand don’t bear thinking about. A positive decision to embrace the technology needs to be taken - the global economy is relying on it. 

N’hésitez pas à vous abonner à notre blog pour connaître les toutes dernières tendances, technologies et actualités.