IS PREDICTIVE SIMULATION A CRYSTAL BALL FOR DATA CENTERS?

This blog was posted on behalf of Jeff Brickley, Director, Data Center Services at DCIM Solutions LLC, and is part-1 of a series of blog posts about predictive simulation for the data center. 

Crystal Ball Draft

What is Predictive Modeling?

By definition Predictive modeling is the process by which a model is created or chosen to try to best predict the probability of an outcome. Most often the event one wants to predict is in the future, but predictive modeling can be applied to any type of unknown event, regardless of when it occurred.

Predictive modeling is used in many industries. For example, the Health Care industry uses historical statistics to create models that predict the likelihood a patient will be readmitted to the hospital.   Other industries take this even further through the use of graphical simulations. This is often referred to as Predictive Simulation or Simulation based performance analytics. Gartner defines Simulation-based performance analytics as: Optimization and simulation using analytical tools and models to maximize business process and decision effectiveness by examining alternative outcomes and scenarios, before, during and after process implementation and execution.

The aerospace industry utilizes Predictive Simulation to test new materials to use in the construction of aircraft that potentially could improve fuel efficiency. Automotive manufacturers use Predictive Simulation to test new parts that will help with turn radius and reduce the likelihood of a roll over. Predictive Simulation is very mature and the physics behind the models has been used for centuries to characterize the behavior of complex systems.

Take the automotive example above. They are essentially using Predictive Simulation to improve performance and reduce risk. Sounds like something we all think about in data centers and IT in general. Wouldn’t it be nice to be able to accurately predict the outcome of any proposed change in your data center? That would be like having a crystal ball that can predict the future. Those don’t exist though; or do they? Predictive Simulation is that crystal ball and is being used to predict the outcome of any proposed change. Many of you may have heard of Computational Fluid Dynamics models of data centers. Computational Fluid Dynamics (CFD) is the physics behind the simulation. There are many data centers in the country today that utilize CFD tools for Predictive Simulation. I personally know of data centers in operation today where you can view the Simulation model in an office and obtain an attribute of a data center such as flow from a floor tile or temperature at a certain point on a rack and then walk to that exact spot in the data center and reality is reflecting exactly what the model predicted. These models are kept up to date in order to run simulations throughout operations to ensure any proposed change will not adversely affect the data center as a whole. Additionally, it is being used to justify capital expenses and at the same time reducing operational costs through efficiency gains.

I have started this blog series in an effort to spark conversation around Predictive Simulation for data centers. I will be discussing the need for data center optimization and then dive into a specific methodology for establishing data center performance goals, utilizing predictive simulation to measure existing performance and then illustrating how predictive simulation can be used ongoing to stay on course to reach those goals. It just so happens that through this process there are significant $$ to be saved and at the same time risk of outage is reduced.

I encourage you to research Predictive Modeling and more importantly Predictive Simulation. You should find that these are not new concepts even to the data center industry. However there may be some rumors out there about how accurate these simulations truly are and whether or not they bring value. My first response to such propaganda is that it all depends on how the models are constructed – garbage in equals garbage out. But make no mistake, a properly calibrated CFD model can be used for Predictive Simulation and thus to accurately predict ANY change in the data center.

About the author: 

1b4f5e3Jeff Brickley is a certified Project Manager and has been in the Information Technology sector for over 17 years.  He has a unique and diverse background including programming, infrastructure and project management.  He has lead large, complex efforts and teams with the common theme throughout his career being that of optimizing operations for business.  His passion for continuous improvement and a background in Mathematics made him an ideal candidate to study, utilize and promote the use of Computational Fluid Dynamics modeling for data centers.

About DCIMdatacenter

http://www.linkedin.com/in/rfschmidt
This entry was posted in Data Center DCIM Datacenter Datacenters Datacentre and tagged , , , , , , , , , , . Bookmark the permalink.

One Response to IS PREDICTIVE SIMULATION A CRYSTAL BALL FOR DATA CENTERS?

  1. Pingback: WHY YOU SHOULD ACE YOUR DATA CENTER | DCIM Data Center Infrastructure and Critical Facility News

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s