What is an IA architecture server?

Generally, an Intel-based server is called an IA (intelarchitecture) architecture server, or a CISC (complexinstructionsetcomputer Complex Instruction Set) architecture server, because the server in the IA architecture is a PC-based x86 processor architecture, the server in the IA architecture is also called a PC server or X86 server. Such as Lenovo's Series servers and HP's NETServer Series servers. (Zdnet Editor note: strictly speaking, the IA Server includes two types, one is the IA-32 server, that is, the common X86 server, the other is the IA-64 server, that is, the anteng server. However, since the current X86 server also began to support 64-bit, IA-32, IA-64 statement is no longer officially adopted by Intel .)

Because the architecture server adopts an open system and features "small, clever, and stable", with reliable performance and low price, in addition, it realizes industrial standardization technology and is supported by a large number of software and hardware suppliers at home and abroad. On the basis of mass production, it is globally based on its extremely high performance and price ratio, it is widely used in China. More file services, printing services, communication services, web services, email services, database services, application services, and other main applications are completed on the Internet and LAN.

Although the IA architecture server starts with a PC, after continuous development, the IA architecture server is far beyond the concept of PC. It differs from PC in the following aspects.

CPU processing capability

Because the server needs to share its data and hardware with the networkProgramProcessing a large amount of data. Therefore, the CPU must have a strong processing capability. Most servers in the IA architecture adopt multi-CPU symmetric processing technology. multiple CPUs perform data operations together, which greatly improves the computing capability of servers and meets the needs of schools in teaching and multimedia applications. In general computer PCs, a single CPU is basically configured, so the data processing capability of PCs is much worse than that of servers. If a PC is used as a server, it often crashes, stops, or starts slowly in daily applications.

I/O (input/output) Performance

In small and medium-sized enterprises or campus network applications, many users often access the server at the same time. There is a large amount of multimedia information transmitted on the network, requiring the I/O (input/output) of the server) strong performance. The server uses scsicard, RAID card, high-speed Nic, memory repeater and other devices, greatly improving the server I/O capabilities. Because PCs are PCs and do not require additional network services, high-performance I/O technology is rarely used on PCs, which is naturally far from the server's I/O performance.

Security and reliability

Because the server is the core device in the network, it must have high reliability and security. The server uses dedicated ECC memory, RAID technology, hot swapping technology, redundant power supply, redundant fan, and other methods to enable the server with fault tolerance and security protection capabilities.

The server must be running continuously for a long time. How long does it take? Different servers have different standards. Generally, the requirement for working group-level servers is that there is no fault during working hours (8 hours a day, 5 days a week; department-level servers require no faults 24 hours a day and 5 days a week. enterprise-level servers have the highest requirements, ensure that the server is ready for use 24 hours a day, 365 days a year. PC is designed for individual users. Therefore, in terms of security and reliability, PC is much lower than server. If a PC is used as a server, it is inevitable that there will be downtime or data loss in daily applications.

Scalability

With the increasing maturity of network information applications, we will inevitably face the problems of expansion and upgrade of network devices. The server has many expansion slots, many drive holders, large hard disks, and memory expansion capabilities, so that the server can meet new requirements during network expansion, this protects device investment costs. As shown in figure 2, the server motherboard has up to eight memory slots and supports up to 16 GB of memory. This scalability is incomparable to PCs.

Manageability

In terms of software and hardware design, servers have sound management capabilities. Most servers are equipped with various sensors on the main board to detect various hardware devices on the server. They can also use corresponding management software to remotely monitor the server, this allows the network administrator to manage the server system in a timely and effective manner. Some management software can remotely detect the signals recorded by sensors on the server board, and remotely monitor and allocate resources to the server. PC does not have a sound hardware management system because of its simple application scenarios. For a lack of professional technical staff, selecting a server with strong manageability can save a lot of trouble.