What are the requirements for the development of memory technology in the era of big data?

With the advent of the era of security big data, the original storage technology of the security industry has been unable to meet the new needs of the industry, especially the construction of public security video surveillance network application has put forward higher requirements for data network sharing, and at the same time "real combat" For the fundamental public security business, big data deep mining relies heavily on data storage systems for the analysis of unstructured data. The emergence of cloud storage technology, in the era of big data development in the security industry, is no different from the revolutionary application, constantly solving the security storage problem, and also providing a powerful driving force for the deep application and development of video surveillance.

In today's world, everyone's words and deeds are generating data and being recorded. The explosive growth of data from all walks of life is driving humanity into the era of big data. According to relevant statistics, the current global data growth rate is about 40% per year. It is estimated that by 2020, the global total data will reach 40ZB.

The data growth is particularly evident in the security industry. In the continuous development and in-depth process of “Safe City”, “Intelligent Transportation” and “Snow Project” in the past two years, the industry development with video surveillance as the core is moving towards super The direction of high-definition, intelligent and converged applications is moving forward, and the data collection of existing video surveillance systems in system engineering is increasing linearly. The emergence of massive data has steadily increased the requirements for efficient and timely storage and processing.

From the current industry point of view, the arrival of the era of big data, video surveillance systems in system engineering have the following main requirements for storage:

First, the monitoring data storage system needs to be scalable, not only to meet the continuous increase of massive data, but also to meet the data needs of collecting higher resolution or more collection points.

Second, the performance requirements of the storage system are high. Different from other fields, video surveillance is mainly the storage of video code stream. In the case of multi-channel concurrent storage, it has high requirements on bandwidth, data capacity, cache, etc., and needs to be optimized for video performance.

Third, big data applications require centralized management analysis of data storage. However, the reality is exactly the opposite. On the one hand, in the process of phased construction of systemic engineering, the purchased equipment cannot be guaranteed to be the same brand. In the actual project, multiple brands and multiple models are everywhere, and the storage for video surveillance is concentrated. Management is very difficult. At the same time, in some large-scale projects, such as the mega-city “Skynet Project”, the road monitoring in the expressway is larger across the region, and centralized storage is more difficult. In addition, due to network bandwidth and old equipment, it is difficult for the system to form a central architecture for unified storage and unified monitoring, resulting in untimely data retrieval in applications.

Fourth, massive data is stored in a timely and efficient manner. According to the current technical regulations and standards, the data acquisition system of the general application field is 7x24 hours, and the time limit for the collection of audio and video information collected by the system is not less than 30 days. The information and some special application areas of audio and video data storage time is longer, even long-term retention, the amount of data increases linearly with time.

Overall, with the in-depth development of systemic security projects and the early emergence of the Internet of Things, the construction of large-scale networked monitoring and the gradual popularization of high-definition surveillance, massive video data has been spurred and impacted on traditional Storage system, unfortunately, the original storage system can not meet the new requirements put forward in the era of big data, the need for new storage technology to support the existing business model, while expanding the new space for artificial intelligence technology in the security field.