New data storage technologies: there is enough space for everyone

Sof Socratous, vice president of Seagate in the EMEA region.

Technologies of the “Internet of Things” promise to give us opportunities that we could only dream of earlier. At the same time, the rapid spread of “smart” devices leads to a no less rapid growth in the volumes of data generated by them.

This raises a number of questions: How to measure and manage this data? What happens to the collected data? What impact do they have on systems and devices? According to Cisco forecasts, by 2020 the number of connected devices worldwide will reach 50 billion units: more than 6.5 devices per user. By 2020, the global amount of data will be about 44 zeb; for comparison, today this figure is only 4 zetabytes. According to IDC forecasts, 13 of these 44 zetabytes will need to be stored somewhere. If we take into account the amount of physical capacity available today, then by 2020 we will be able to store only about 6.5 Zetabytes of data. This means that in order to translate the concept of the “Internet of things” into reality, a qualitative leap in storage technologies and service solutions is necessary.


')
Today, several innovative technologies are being developed, which in the long term can help solve this problem. For example, the technology of thermo-assisted magnetic recording (HAMR) can potentially increase the density of magnetic recording by a factor of 100. The problem is that it is unlikely to be available to commercial users until 2016. Therefore, in anticipation of this blessed day, we will have to solve the problem of increasing the efficiency of data storage. Not so long ago, we suggested putting companies and employees on an “informational diet” so that they consume less data. Obviously, the availability of competent data management policies, coupled with deduplication technology and reliable backup strategies will help to more effectively organize data storage. Of course, this alone is not enough, but, fortunately, there are developments for clouds and data centers. Storing a large amount of data is quite expensive, so another logical step would be to create storages that require less capital and operating costs.

Today there are several similar technologies. For example, the Kinetic platform allows you to make hard drives smarter. They can exchange data with the entire storage pool and together make “decisions” about where and how to save data. This allows you to reduce the cost of ownership by 40%, increase the flexibility and performance of storage, as well as increase the density of drives in the rack. Another approach that has been gaining popularity lately is the use of a multi-level data storage model. By placing conventional hard drives, solid-state hybrid (SSHD) and solid-state (SSD) drives in an array, IT departments can more efficiently organize data. For example, they can place the most important and frequently used information on solid-state devices, and the one that is used less often on conventional HDDs.



In the future, there will be even more data centers using this approach. Much work remains to be done to make the Internet of Things a reality, but already today there are a number of developments that inspire optimism. The main thing is to understand the existing problems and continue to invest in the creation of new technologies. This is the challenge facing the IT industry as a whole, so we must work together. Only then can we bring the concept of universal interconnectedness to life.