What we think is going to happen is that about 40% of the data generated by IoT will have to be processed at the...
edge, simply because the information has to be turned around pretty quickly.
So, two things can happen: One is you can create micro data centers at the edge of the network. Because you have to have software and applications, business analytics -- all the things you have at the enterprise data center -- that could be one scenario. The other scenario is that this data goes back to the enterprise. Now, if it goes back to the enterprise, there's a probability that you're going to have to have a cloud infrastructure to be able to support this. The reason is fairly logical. In 2020 there will be 30 billion devices that will be connected to the network. At that point we think that every 40 minutes there'll be 400,000 devices connected. That's a lot of devices being connected. If you're an IT manager, there's no way you can use normal IT infrastructure; you have to use a cloud infrastructure that scales really well.
Also, instead of bring your own device, which stressed out systems management because you're bringing in all these devices, now with the IoT we're getting devices that don't look like anything IT traditionally had to support. So, the advice is, first of all, beef up your systems management process and procedures to enable the variety and the frequency of devices to be brought on to your network. That's No. 1. No. 2 is think of the network bandwidth and what that means for you over the next three years. Today you probably have enough capacity to bring on these devices, but without knowing the characteristics of those devices, your network probably will be stressed. Wacky things happen -- devices get out of control, they generate too much data -- the next thing is your network's flooded.
The third thing is with respect to the service providers. I think you have to be prepared to put more corporate data on the service provider platform. The big thing for us is that we think the IoT will force traditional enterprise data centers or data center owners to move their data onto a service provider platform simply because you want to do what we call IoT data blending. Data blending means taking the data, finding the value and using it to benefit the business.
About the author:
Vernon Turner is the senior vice president of IDC's enterprise infrastructure, mobility, consumer, network, telecom, Internet of Things, and sustainability research. In addition to driving research on the evolution of the Internet of Things, he is a member of the IoT World Forum Steering Committee.
For a detailed summary of IDC's outlook on the IoT, check out this blog post: IDC prediction for IoT 2015: It's a doozy.
Should CIOs care that IBM plans to spend $3 billion on an IoT business unit?
IoT data security and privacy prioritization needs to happen now
Dig Deeper on Internet of Things (IoT) Network Infrastructure
Have a question for an expert?
Please add a title for your question
Get answers from a TechTarget expert on whatever's puzzling you.