Introduction
The main frequency of the CPU, that is, the clock frequency of the CPU core (CPU Clock Speed). How many megahertz does the so-and-so CPU usually refer to, and this number of megahertz is the "CPU's main frequency". Many people think that the main frequency of the CPU is its operating speed, but it is not. The main frequency of the CPU indicates the oscillation speed of the digital pulse signal in the CPU, and has no direct relationship with the actual computing power of the CPU (that is to say, the current CPU frequency does not directly affect the computing power of the CPU. Impact. It's just because the CPU frequency is lower now, it is much higher than other hardware such as memory).
For example, most of AMD’s AthlonXP series CPUs can achieve the higher CPU performance of Intel’s Pentium 4 series CPUs at a lower frequency, so AthlonXP series CPUs use the PR value. To name it. Therefore, the main frequency is only an aspect of CPU performance, and does not represent the overall performance of the CPU. The main frequency of the CPU does not represent the speed of the CPU, but increasing the main frequency is essential to increase the CPU's computing speed. For example, if a certain CPU executes an arithmetic instruction in one clock cycle, when the CPU runs at 100MHz, it will be twice as fast as when it runs at 50MHz. Because the 100MHz clock cycle takes half the time compared to the 50MHz clock cycle, that is, the time required for a CPU working at 100MHz to execute an operation instruction is only 10ns shorter than 20ns when working at 50MHz, which is natural operation. The speed is also doubled. It's just that the overall operating speed of the computer depends not only on the computing speed of the CPU, but also on the operating conditions of other sub-systems. Only when the main frequency is increased, the operating speed of each sub-system and the data transmission speed between each sub-system can be After being improved, the overall operating speed of the computer can be truly improved.
Features
There is a certain relationship between the main frequency and the actual computing speed, but there is no definite formula that can quantify the numerical relationship between the two, because the computing speed of the CPU depends on Performance indicators of all aspects of the CPU's pipeline (cache, instruction set, CPU bits, etc.). Since the main frequency does not directly represent the calculation speed, under certain circumstances, the actual calculation speed of the CPU with a higher main frequency is likely to be lower. For example, most of AMD’s AthlonFX series CPUs can reach the higher frequency of Intel’s Pentium 4 series CPUs at lower frequencies, so AthlonFX series CPUs are named after the PR value. Therefore, the main frequency is only an aspect of CPU performance, and does not represent the overall performance of the CPU.
The main frequency of the CPU does not represent the speed of the CPU, but increasing the main frequency is crucial to increasing the CPU's computing speed. For example, suppose a certain CPU executes an arithmetic instruction in one clock cycle, then when the CPU runs at 100MHz, it will be twice as fast as it runs at 50MHz. Because the 100MHz clock cycle takes half the time compared to the 50MHz clock cycle, that is, the time required for a CPU working at 100MHz to execute an operation instruction is only 10ns shorter than 20ns when working at 50MHz, which is natural operation. The speed is also doubled. It's just that the overall operating speed of the computer depends not only on the computing speed of the CPU, but also on the operating conditions of other sub-systems. Only when the main frequency is increased, the operating speed of each sub-system and the data transmission speed between each sub-system can be After being improved, the overall operating speed of the computer can be truly improved.
Improving the CPU operating frequency is mainly restricted by the production process. Because the CPU is manufactured on a semiconductor silicon chip, wires are required to connect between the components on the silicon chip. Because the wires are required to be as thin as possible under high frequency conditions, so as to reduce stray interference such as wire distributed capacitance In order to ensure the correct operation of the CPU. Therefore, the limitation of the manufacturing process is one of the biggest obstacles to the development of CPU frequency.
When it comes to processor clock speed, we must mention two closely related concepts: multiplication and FSB. FSB is the reference frequency of the CPU, and the unit is also MHz. The FSB is the speed at which the CPU and the motherboard run synchronously, and in most computer systems, the FSB is also the speed at which the memory and the motherboard run synchronously. In this way, it can be understood that the FSB of the CPU is directly connected to the memory. Through, realize the synchronous operation state between the two; the frequency multiplier is the multiple of the ratio of the main frequency to the external frequency. Main frequency, FSB, and frequency multiplier, the relationship is: main frequency = FSB × frequency multiplier. Early CPUs did not have the concept of "frequency multiplication". At that time, the main frequency and the speed of the system bus were the same. With the development of technology, the CPU speed is getting faster and faster, and the accessories such as memory and hard disk are gradually unable to keep up with the speed of the CPU. The emergence of frequency multiplier solves this problem. It can make the memory and other components still work at a relatively low level. Under the system bus frequency, the main frequency of the CPU can be increased infinitely by multiplying (theoretically). We can regard the FSB as a production line in the machine, and the frequency multiplier is the number of production lines. The production speed of a machine (main frequency) is naturally the speed of the production line (FSB) multiplied by the number of production lines. (Multiplier) too. Manufacturers have basically locked the frequency multiplier. The only way to overclock is to start with the FSB. Through the combination of multiplier and FSB, jumper the motherboard or set soft overclocking in the BIOS, so as to achieve a partial improvement in the overall performance of the computer. So pay attention to the FSB of the CPU as much as possible when buying.
How to view
1. In Windows, right-click the "My Computer" icon on the desktop and select "Properties" to view. In the mac system, click the apple icon in the upper left corner of the screen and select the first item (About This Mac) to view it.
2. Press pause break when booting. At this time, because it is the system power-on self-test, you can view the CPU frequency in the BIOS.
3. Use CrystalCPUID software to view. This is a processor information detection overclocking tool, which has basically the same function as WCPUID, but CrystalCPUID supports a wider range of processors. CrystalCPUID supports almost all types of processor detection, the most special is that CrystalCPUID has complete processor and system information.
Restrictions
Improving the CPU frequency is mainly restricted by the production process. Because the CPU is manufactured on a semiconductor silicon chip, wires are required to connect between the components on the silicon chip. Because the wires are required to be as thin as possible under high frequency conditions, so as to reduce stray interference such as wire distributed capacitance In order to ensure the correct operation of the CPU. Therefore, the limitation of the manufacturing process is one of the biggest obstacles to the development of CPU frequency.
Frequency and speed
The relationship between frequency and speed: Generally speaking, the number of instructions completed in a clock cycle is fixed, so the higher the frequency, the higher the speed of the CPU. Coming soon. However, because the internal structure of various CPUs is not the same, the main frequency cannot be used to summarize the performance of the CPU. But the CPU frequency can determine the grade and price level of the computer. Take Pentium 4 2.0 as an example. Its main frequency is 2.0GHz. What does this show?
Specifically, 2.0GHz means that it will generate 2 billion clock pulse signals per second, and each clock signal period is 0.5 nanoseconds. The Pentium 4 CPU has 4 pipeline operation units. If the load is even, the CPU can perform 4 binary addition operations in 1 clock cycle.
This means that the Pentium 4 CPU can perform 8 billion binary addition operations per second. But such an amazing computing speed cannot fully serve users, and the computer hardware and operating system itself consume CPU resources. However, the Athlon XP processor adopts the PR nominal method. The conversion calculation formula between the nominal frequency and the actual frequency of the Athlon XP processor with a 266MHz front-side bus frequency disclosed by AMD is as follows: nominal frequency=3×actual frequency/2-500 actual frequency =2×nominal frequency/3+333 For example, the actual frequency of Athlon XP 2100+ is 1733MHz=2×2100/3+333.
Memory clock speed
The memory clock speed is the same as the CPU clock speed. It is customarily used to indicate the speed of the memory, and it represents the highest operating frequency that the memory can achieve. The main frequency of the memory is measured in MHz (megahertz). The higher the memory clock speed, to a certain extent, the faster the memory can reach. The main frequency of the memory determines the maximum frequency at which the memory can work normally.
The clock speed of a computer system is measured by frequency. The crystal oscillator controls the clock speed. When voltage is applied to the quartz wafer, it vibrates in the form of a sine wave. This vibration can be recorded by the deformation and size of the wafer. The vibration of the crystal is expressed in the form of a sinusoidal harmonic and a changing current, and this changing current is the clock signal. The memory itself does not have a crystal oscillator, so the clock signal when the memory is working is provided by the north bridge of the motherboard chipset or directly by the clock generator of the motherboard, which means that the memory cannot determine its own operating frequency, its actual operating frequency It is determined by the motherboard.
The frequency of DDR memory and DDR2 memory and DDR3 memory can be expressed in two ways: working frequency and equivalent frequency. The working frequency is the actual working frequency of the memory particles, but because the DDR memory can rise in the pulse And the falling edge transmit data, so the equivalent frequency of transmitting data is twice the operating frequency; while each clock of DDR2 memory and DDR3 memory can read/write data at four times the operating frequency, so the equivalent of transmitting data The frequency is four times the operating frequency. For example, the operating frequencies of DDR 200/266/333/400 are 100/133/166/200 MHz, and the equivalent frequencies are 200/266/333/400 MHz; the operating frequencies of DDR2 400/533/667/800 are respectively It is 100/133/166/200 MHz, and the equivalent frequency is 400/533/667/800 MHz; the working frequency of DDR3 1066/1333/1600/1800/2000 is 266/333/400/450/500 MHZ respectively , And the equivalent frequencies are 1066/1333/1600/1800/2000MHZ.
The memory asynchronous working mode has many meanings. In a broad sense, any memory working frequency is inconsistent with the CPU's FSB can be called the memory asynchronous working mode. First of all, the earliest memory asynchronous working mode appeared in the early motherboard chipset, which can make the memory work in a mode that is 33MHz higher or 33MHz lower than the CPU FSB (note that it is only a simple difference of 33MHz), which can improve system memory performance or use The old memory continues to exert its waste heat. Secondly, in normal working mode (CPU is not overclocked), many motherboard chipsets also support asynchronous memory working mode. For example, Intel 910GL chipset only supports 533MHz FSB or 133MHz CPU FSB, but it can match the working frequency. DDR 266 with 133MHz, DDR 333 with a working frequency of 166MHz, and DDR 400 with a working frequency of 200MHz work normally. There are differences in memory performance. Thirdly, in the case of CPU overclocking, in order not to make the memory drag the CPU overclocking ability, the working frequency of the memory can be lowered to facilitate overclocking. For example, AMD’s Opteron 144 with Socket 939 interface is very easy to overclock. Many products The FSB can easily exceed 300MHz. If in the memory synchronization mode, the equivalent frequency of the memory will be as high as DDR 600. This is obviously impossible. In order to smoothly exceed the 300MHz FSB, we can Before overclocking, set the memory to DDR 333 or DDR 266 in the motherboard BIOS. After over 300MHz FSB, the former is only DDR 500 (some high-quality memory can reach), and the latter is only DDR 400 (completely Normal standard frequency), it can be seen that the correct setting of the asynchronous mode of the memory will help the overclocking success.
DDR4 memory is a new generation of memory specifications. On January 4, 2011, Samsung Electronics completed the first DDR4 memory in its history.
There are three biggest differences between DDR4 and DDR3: 16bit prefetch mechanism (DDR3 is 8bit), the theoretical speed is twice that of DDR3 at the same core frequency; more reliable transmission specifications, data reliability is further improved ; The working voltage drops to 1.2V, which is more energy-efficient. Strictly speaking, DDR4 should be called DDR4 SDRAM. The full name of DDR4 SDRAM is Double Data Rate Fourth Synchronous Dynamic Random Access Memory, which is the fourth generation of double data rate synchronous dynamic random access memory.
It wasn't until 2014 that DDR4 memory was used for the first time. The first Intel flagship x99 platform to support DDR4 memory was Intel’s flagship x99 platform. Advantages, but if users wanted to experience the flagship platform, they could only buy high-priced DDR4, because x99 only supports DDR4.
Almost all motherboard chipsets support asynchronous memory. Intel supports it from the 810 series to the newer 875 series, while VIA provides this feature from the 693 chipset.
In 2018, Intel released the KBL-G processor belonging to the eighth-generation Core processor family, which is the long-rumored Intel CPU + AMD GPU processor. There are 5 types of KBL-G processors, the highest specification is i7-8809G, 4 core 8 threads, main frequency 3.1 GHz, maximum turbo frequency 4.2 GHz, 8 M L3 cache, memory supports dual channel DDR4-2 400 MHz, And does not lock the multiplier, the graphics card is equipped with Radeon RX Vega M GH.
Comparison with FSB
The main frequency of the CPU is the clock frequency of the CPU core (CPU Clock Speed). The main frequency of the CPU does not represent the speed of the CPU, but increasing the main frequency is very important to increase the computing speed of the CPU. Assuming that a certain CPU executes an arithmetic instruction in one clock cycle, when the CPU runs at a frequency of 100MHz, it will be twice as fast as when it runs at a frequency of 50MHz. However, the overall operating speed of the computer depends not only on the computing speed of the CPU, but also on the operating conditions of other sub-systems.
The external frequency is the reference frequency of the CPU and the entire computer system, and the unit is MHz. In early computers, the synchronous operation speed between the memory and the motherboard was equal to the FSB. In this way, it can be understood that the external frequency of the CPU is directly connected to the memory to achieve a synchronized operation state between the two. For the current computer system, the two can be completely different, but the meaning of FSB still exists. Most frequencies in the computer system are realized by multiplying the FSB by a certain multiple.
Solution to the deviation of CPU clock speed
Fault phenomenon:
The CPU is Core2DuoE6300, and the latest version of EVEREST measured the CPU clock speed as 12058MHz. The multiplier is 6×, the external bus frequency of the CPU is 201.0MHz, the memory bus frequency is 2680MHz, and the DRAM:FSB ratio is 8:6. Why is there so much difference between the existing operating frequency and the actual operating frequency? Is there a problem with the external bus frequency setting of the CPU?
Analysis and processing:
The reason why the CPU is currently running The large difference between the frequency and the actual operating frequency is related to the improper setting of the CPU parameters in the BIOS. According to the different motherboard specifications, you can enter the "CPU voltage and frequency" related setting options, adjust the FSB frequency to 266MHz, adjust the multiplier to 7, and select the automatic setting option for the memory frequency. After setting and saving, the CPU frequency will be It will become 1.86GHz (266MHzX7).