According to a Cisco survey, some 50 billion devices will be connected to the Internet by 2020 – from hall-sized, computer-controlled fermentation tanks at industrial bakeries to quasi-intelligent doorknobs. All of these devices produce data and will overtake humans as the biggest producers of data in the not-too-distant future. For much of this data, the Cloud is the right partner. In view of the fact that many IoT devices are small, need to run on a tiny amount of electric power and perform only narrowly defined functions, they need a “big brother” who can give them the “right” amount of computing power. This is a job which the Cloud performs perfectly. But in the case where information is needed only locally and for just a short period of time, the situation becomes quite different: Data of this kind needs to be processed quickly, preferably onsite, without any need to store the data.
In cases like these, Fog Computing represents the ideal solution. Previously known as “Edge Computing”, this technology processes data decentrally, on the edges of the network. Due to major improvements in the computing power and power consumption of processors over the past few years, tiny computers are now capable of carrying out some highly complex calculations. One possible Fog Computing scenario involves future car-2-car communication in emergencies. If a vehicle slams on its brakes on the freeway, the other vehicles within a radius of several hundred yards need to be informed of this to prevent collisions. To send this acceleration data via the Cloud to a data center, which would determine all the relevant vehicles and communicate the required information back to them would be too time-consuming. Here the necessary data processing and the resultant communication with nearby vehicles needs to take place using local computing power.
Regardless of the actual application cases, Fog Computing can be viewed as an additional virtualized layer between the data producer and the Cloud. It represents a local decision-making and processing level which can relieve the pressure on other levels further downstream. This layer prepares data according to established rules, thus improving response times while reducing the broadband requirements for potential Cloud connections as well as the required storage capacity at data centers.
CeBIT 2016 will feature several display categories illustrating how data centers can cope with these new challenges and highlighting the rapid evolution of Cloud technology. DatacenterDynamics and CeBIT for instance will be merging their competence again in Hall 12, where visitors will find international manufacturers, service providers, operators and suppliers of IT infrastructure as well as a high-caliber conference with no attendance fee. Hall 13 on the other hand will serve as the first port of call for the Internet of Things, while the neighboring topic clusters of Communication & Networks and I oT SOLUTIONS will turn the hall into a unique platform with vast synergy potential.
Fog Computing facilitates applications whose insufficient bandwidth means it is impossible for them to rely solely on the Cloud or which produce data which are important only onsite, in a local context, and therefore do not require any central processing. The more IoT devices populating our factories, streets and buildings, the bigger the role of Fog Computing will become in insuring an efficient and affordable Internet of Things.