From different angles to observe the server, there can be different classification of the server, the following we discuss the classification of the server from different angles.

This classification is mainly based on the different processor architectures used by the two servers.RISC architecture server USES the CPU is the so-called reduced instruction set processor, reduced instruction set is the main characteristic of CPU USES fixed-length instructions, using assembly line executes instructions, the processing of such an instruction can be divided into several stages, processing unit executes instructions that the processor set different at different stages, such as order processing if divided into three phases,When the NTH instruction is in the third processing stage, the N+1 instruction will be in the second processing stage, and the N+2 instruction will be in one processing stage.This pipelining of instructions gives the CPU the ability to process instructions in parallel, which allows the processor to process more instructions per unit of time.IA architecture server USES a CISC architecture, namely complex instruction set architecture, the architecture is the feature of instruction is longer, the function of the instruction is stronger, a single command executable function is more, so that we can increase the computing unit, make the function of an instruction execution can be executed in parallel at the same time to improve the operation ability.Both architectures have been growing in competition with each other for a long time, and both have grown rapidly.The server of IA architecture adopts the open architecture, thus has a large number of hardware and software supporters, and has had considerable development in recent years.

This classification method is a relatively old classification method, main is classified according to the size of the application server environment, such as a 10 around the client computer network environment is suitable for use a workgroup server, the server often USES one processor, hard disk capacity of small and not strong network capacity;A few client computer network for departmental servers, departmental server relative ability stronger, often using 2 processor, memory and disk capacity, disk I/O and network I/O capability is strong, so that this server can have enough capacity to accept service requirements put forward by the client;And enterprise servers are often more than one hundred clients in the network environment, in order to take a lot of the response of the service request, the server often USES four processors, there is a lot of hard disk and memory, and can be further extended to meet the higher demand, because to deal with a lot of access at the same time, so the server disk and network speed should be high.In order to achieve this requirement, it is often necessary to use multiple network cards and multiple hard disk parallel processing.All of the above description is not accurate, there are a lot of special circumstances, such as a network client may be many, but access to the server may have little, there is no need to a strong enterprise server function, due to the presence of these elements, so that the classification of the this server methods tend to be more qualitative, not quantitative,That is to say, the performance of the server is gradually increasing from the team level to the department level to the enterprise level, as are various other features.

For example, file/print servers, a type of server that performs file storage and printer resource sharing services, are still widely used in office environments today.Database server, which runs a database system for storing and manipulating data and providing data query and modification services to networked users, is also a server widely used in business systems.Web server, E-mail server, NEWS server, PROXY server, these servers are typical of Internet applications, they can complete the home page storage and transmission, E-mail service, newsgroup service, etc.All of these servers are more than just hardware systems, they often use a combination of hardware and software to perform their specific functions.