6 | THE DATA CENTER JOURNAL
www.datacenterjournal.com
Machine to machine
• Sensors to controlled devices
• Networked computers
Many of the forms of data exchange
above have predictable patterns and pre-
dictable scaling. Two of the forms that are
extremely difficult to predict with rapid
scaling are social media and machine
to machine where machine learning is
involved. Almost every social media
company that has boomed has done so
unexpectedly and very quickly, outpacing
any and all predictions.
Social media data generation and
communication is still undergoing
explosive growth. At the same time the
storage and processing (analysis) of this
data is undergoing even more significant
growth. e use and monetization of this
information is only just beginning to be
understood.
Social media data centers sub-
sequently require a different scaling
approach than many other types of data
centers. Not only do they have to account
for the accelerating generation and com-
munication of this data, they also have to
account for the long-term storage and pro-
cessing. e combination of these requires
The ConversaTion Prism
Brought to you by Brian Solis & JESS3
Figure 1