At present, the server CPU is still according to the CPU instruction system to distinguish, usually divided into two categories of CISC-type CPU and RISC-type CPU, and later appeared a 64-bit VLIM (Very Long Instruction Word ultra-long instruction set architecture) instruction system CPU.
1.
CISC is the abbreviation of English "Complex Instruction Set Computer", which means "Complex Instruction Set" in Chinese, and refers to the x86 (a naming convention of Intel CPU) series of CPUs produced by Intel. It refers to the x86 (a naming convention of Intel CPUs) series of CPUs produced by Intel and its compatible CPUs (CPUs produced by other vendors such as AMD, VIA, etc.), which are based on the PC (personal computer) architecture. This kind of CPU is generally a 32-bit structure, so we also make it IA-32 CPU.(IA: Intel Architecture, Intel Architecture).CISC type CPU currently has two main categories: intel's server CPUs and AMD's server CPUs.
2, RISC-type CPU
RISC is the English "Reduced Instruction Set Computing " abbreviation, the Chinese meaning is "streamlined instruction set". It is in the CISC (Complex Instruction Set Computer) instruction system on the basis of the development of some people on the CISC machine test shows that the frequency of use of a variety of instructions is quite disparate, the most commonly used are some of the simpler instructions, which accounted for only 20% of the total number of instructions, but in the program frequency but accounted for 80%. Complex instruction systems inevitably increase the complexity of the microprocessor, making the processor's development time is long and high cost. And complex instructions require complex operations, which will inevitably reduce the speed of the computer.
Based on the above reasons, in the 1980s, RISC-type CPUs were born. Compared to CISC-type CPUs, RISC-type CPUs not only streamline the instruction system, but also use a structure called "superscalar and super pipeline", which greatly increases the parallel processing capability (parallel processing parallel processing is). Parallel processing means that a server has multiple CPUs processing at the same time. Parallel processing is the simultaneous processing of multiple CPUs on a server. Parallel processing can greatly increase the data processing power of a server. (Departmental and enterprise-level servers should support CPU parallel processing technology).
That is to say, the architecture at the same frequency, using RISC architecture CPU than CISC architecture CPU performance is much higher, which is determined by the technical characteristics of the CPU. The RISC instruction system is more suitable for UNIX, the operating system of high-end servers, and now Linux also belongs to the UNIX-like operating system. The RISC-type CPUs are not compatible with Intel and AMD CPUs in either software or hardware.
Currently, the CPUs that use RISC instructions in mid-range and high-end servers are mainly of the following types: PowerPC processors, SPARC processors, PA-RISC processors, )MIPS processors, and Alpha processors.
From the current state of server development, the IA architecture (CISC architecture) PC servers characterized by "small, smart, stable" have been more widely used because of their reliable performance and low price. In the field of Internet and LAN, for file services, printing services, communication services, Web services, e-mail services, database services, application services and other purposes.
The last point worth noting is that although the CPU is one of the most important factors in determining server performance, the CPU can't perform as well as it should without the support and cooperation of other accessories.
Processor main frequency
The main frequency, that is, the clock frequency of the CPU, is simply an abbreviation of the operating frequency of the CPU (the number of synchronized pulses occurring in one second). The unit is Hz. It determines the operating speed of the computer, with the development of computers, the main frequency from the past MHZ development to the current GHZ (1G = 1024M). Generally speaking, in the same series of microprocessors, the higher the main frequency on behalf of the computer speed is also faster, but with different types of processors, it can only be used as a parameter for reference. In addition, the CPU's computing speed also depends on the CPU pipeline of various aspects of performance indicators. Since the main frequency does not directly represent the computing speed, so in some cases, it is likely that the CPU with a higher main frequency will actually have a lower computing speed. Therefore, the main frequency is only one aspect of CPU performance, and does not represent the overall performance of the CPU.
Speaking of processor frequency, it is necessary to mention two concepts closely related to it: multiplier and external frequency, external frequency is the benchmark frequency of the CPU, the unit is also MHz. external frequency is the speed of synchronous operation between the CPU and the motherboard, and the vast majority of the current computer system in the external frequency is the synchronous operation of the speed of synchronous operation between the memory and the motherboard, in this way, it can be understood as the external frequency of the CPU directly connected to the memory, to achieve between them, the external frequency is the same as the speed of the motherboard. In this way, it can be understood that the external frequency of the CPU is directly connected to the memory to realize the synchronous operation between the two; the multiplier frequency is the multiple of the ratio between the main frequency and the external frequency. The main frequency, external frequency, multiplier frequency, its relationship formula: main frequency = external frequency × multiplier frequency.
Early CPUs did not have the concept of "multiplier", when the main frequency and the system bus speed is the same. With the development of technology, CPU speed is getting faster and faster, memory, hard disk and other accessories gradually can not keep up with the speed of the CPU, and the emergence of the multiplier to solve this problem, it can make the memory and other components are still working in a relatively low system bus frequency, while the main frequency of the CPU can be multiplied to unlimitedly improve (theoretically).
We can think of the external frequency as a production line within the machine, and the multiplier as the number of lines in the production line. The speed of a machine's production (the main frequency) is naturally the speed of the production line (the external frequency) multiplied by the number of lines in the production line (the multiplier). Now manufacturers have basically locked the multiplier, to overclocking only from the external frequency to start, through the multiplier and the external frequency with the motherboard's jumpers or in the BIOS set up a soft overclocking, so as to achieve the overall performance of the computer part of the enhancement. So try to pay attention to the external frequency of the CPU when you buy it.
Processor external frequency
The external frequency is the base frequency of the CPU and even the entire computer system, in MHz (megahertz). In early computers, the synchronization between the memory and the motherboard ran at a speed equal to the external frequency, and in this way, it can be understood that the external frequency of the CPU is directly connected to the memory to achieve a synchronous running state between the two. For the current computer system, the two can be completely different, but the significance of the external frequency still exists, most of the frequency in the computer system is based on the external frequency, multiplied by a certain number of times to achieve, this multiplier can be greater than 1, can also be less than 1.
Speaking of processor external frequency, we have to mention two concepts that are closely related: the multiplier frequency and the main frequency, the main frequency is the clock frequency of the CPU; the multiplier frequency that is the main frequency and the external frequency of the ratio of the multiplier. The main frequency, external frequency, multiplier frequency, its relationship formula: main frequency = external frequency × multiplier frequency.
Before 486, the main frequency of the CPU is still in a lower stage, the main frequency of the CPU is generally equal to the external frequency. And after the appearance of 486, because the CPU operating frequency continues to increase, while some other devices of the PC (such as plug-in cards, hard disk, etc.) but by the process limitations, can not withstand a higher frequency, thus limiting the further increase in the frequency of the CPU. Therefore, the emergence of multiplier technology, the technology can make the CPU internal operating frequency into a multiple of the external frequency, so as to enhance the multiplier frequency to achieve the purpose of raising the main frequency. This means that external devices can operate at a lower external frequency, and the main CPU frequency is a multiple of the external frequency.
In the Pentium era, the external frequency of the CPU was generally 60/66MHz, and from the Pentium Ⅱ 350 onwards, the external frequency of the CPU was increased to 100MHz, and currently the external frequency of the CPU has reached 200MHz. The newest addition to the CPU is a new, more powerful, and more efficient way to improve the overall performance of your computer.
The external frequency is easily confused with the front-side bus (FSB) frequency. The FSB speed refers to the speed of the bus between the CPU and the Northbridge chip, and more substantially represents the speed of data transfer between the CPU and the outside world. The concept of external frequency is based on the digital pulse signal oscillation speed, that is to say, 100MHz external frequency refers to the digital pulse signal oscillation 10,000,000 times per second, which affects the frequency of the PIC and other buses.
The main reason why the two concepts of Front Side Bus and External Frequency are easily confused is that in the previous long period of time (mainly before and just after Pentium 4), the Front Side Bus frequency and External Frequency are the same, and therefore tend to directly refer to the Front Side Bus as the External Frequency, which ultimately leads to such a misunderstanding.
As computer technology evolved, it was realized that the front bus frequency needed to be higher than the external frequency, so QDR (Quad Date Rate) technology, or other similar technologies, were used to achieve this currently. The principle of these techniques is similar to AGP's 2X or 4X, which makes the front bus frequency 2X, 4X or even higher than the external frequency, and since then the difference between the front bus and the external frequency has only begun to be emphasized.