Currently, a person uses many different devices and computer systems. One of the everyday computer systems that people use is the personal computer. Data on personal computers is collected using the Windows operating program on which the system is running. By default, Windows collects “full” diagnostic data and sends it to Microsoft, a big plus for storing information. However, the downsides are that the system is prone to data leakage and the collection of a considerable amount of user data, which can overload the storage. An indisputable plus is recovering all their data if they are lost for the user. However, a significant disadvantage is the lack of privacy for the user.
In addition, there is such a type of computer system as the mainframe. It is a universal high-performance server with a large RAM and external memory. The mainframe collects user data on a direct access storage device or optical media. Thus, data can be retrieved directly and sequentially, which is convenient depending on the task. Application programs do not need to collect initial information from multiple sources by storing it on a single server. However, a significant disadvantage for users is the user interface. The mainframe has weak communication between the user and the computer system. However, it is now possible to provide a web interface at a minimal cost.
Moreover, a computing cluster (supercomputer) consists of computing nodes united by a high-speed communication network. Each computing node has its RAM and solid-state drive and is connected to shared parallel network storage (Jouppi et al., 2020). The advantage is that, following the rules, the user independently provides a backup of his data. However, a significant disadvantage is that data storage is no longer involved in calculations on a supercomputer is not allowed. The maximum amount of data for a user is determined by the disk quota determined during registration, while in the future, the percentage can be increased if necessary.
Reference
Jouppi, N. P., Yoon, D. H., Kurian, G., Li, S., Patil, N., Laudon, J.,… & Patterson, D. (2020). A domain-specific supercomputer for training deep neural networks. Communications of the ACM, 63(7), 67-78.