Five Reasons your Data Center’s (ACE) Availability, Capacity & Efficiency are being Compromised
Data center owner-operators are increasingly looking for solutions to minimize total cost of ownership, cost per kW of IT load, and downtime. This paper explains the five main
contributors to runaway data center costs, then introduces the ACE performance score and the continuous modeling process. Using both, this paper briefly explains how they are helping owner-operators save millions of dollars annually per data hall.
Could ‘minimize’ be the verb that best sums a data center owner-operator’s
ultimate objective?
Think about it, whatever business you’re in, and whichever type of data center(s)
you own, you almost certainly want to minimize one or more of the following:
• Cost overruns
• TCO (total cost of ownership)
• Cost per kilowatt ($/kW) of IT load
• Downtime
In an industry where the average TCO overspend is around $27m per MW, where $/kW can spiral out of control within just a few short years of entering operation, and where the average cost of downtime is $627k per incident, owner-operators want solutions.
Poor planning and inefficient use of power, cooling and space represents a
significant threat to your efforts to minimize costs. Yet it is precisely this that so
often forces you into a corner – build a new facility to take the strain, or invest in
a major overhaul. Neither ‘solution’ is attractive, so why are owner-operators so
frequently in a position where their aspirations are never realized?
In this paper, we set out to not only answer that question, but to also offer a
solution going forward.
First, we identify the five major contributors to increased costs and downtime. Then we propose that the greatest opportunity to minimize these can be achieved by adopting a simple, inexpensive solution: the ACE performance score.
The ACE performance score is unique way of assessing and visualizing the three
critical indicators of data center performance, as described below. It works by
mapping data from DCIM toolsets into a powerful 3D Virtual Facility model. With
that automated process accomplished, it simulates the resulting distribution of
airflow and temperature in the space. This confluence of predictive modeling and
DCIM data is called Predictive Modeling for DCIM.
The ACE performance score can be used from inception through operation, and
it considers the dynamic interrelationship of the three variables – ACE – that
ultimately dictate how well a data center performs and, by extension, how costly
it is to run:
• Availability (A) of IT, including during power and cooling failures
• How much capacity (C) is available to install, power and cool additional IT
• How efficient (E) the cooling delivery is to the IT
With the ACE performance score introduced and explained, we conclude by
introducing a simple business process through which ACE can be easily applied:
continuous modeling.
