In a narrow sense, overclocking is to increase the operating frequency of the CPU to improve the performance of the entire machine. Broadly speaking, any behavior and related actions that can increase the operating frequency of a certain part of the computer can be called overclocking. When referring to overclocking in this book, if not stated otherwise, the latter definition shall prevail. The origin of overclocking cannot be verified at present, and no one knows who is the initiator, but its development process can still be traced. ?
The emergence of overclocking is actually a loophole in the production process of CPU manufacturers, and it all starts with the special production process of the CPU. CPU is a high-tech crystallization and can represent the latest human technology, so its manufacturing also requires the most advanced technical support. Precisely because CPUs are always at the forefront of technological development, even with the strength of companies like Intel, they are unable to completely monitor the CPU production process, which means that there are uncontrollable factors involved. This creates a more serious problem - the final working frequency of the CPU cannot be completely determined. To put it simply, when a certain production line manufactures a specific model of CPU, it can only ensure that the final product is within a certain frequency range, and it is impossible to "exactly" set it at a required frequency. As for how serious the deviation is, it depends on the specific production process level. For example, AMD's old production line had to be modified because the yield rate was too low. Of course, the CPU produced in this way cannot be put into the market with the same main frequency, so it has to be marked according to the actual frequency after it is offline. However, as a manufacturer, there is undoubtedly a certain risk if the CPU is directly marked with the measured frequency, so a safer approach is to mark the CPU one or two levels lower to ensure reliability and stability. ?
The little leeway reserved by manufacturers to ensure product quality became the initial source of overclocking inspiration for a small number of super hardware enthusiasts, and they decided to regain this lost performance. After unremitting research, some geniuses finally realized that the actual working frequency of the CPU is completely determined by several jumpers on the motherboard, so the initial overclocking began from then on.
The working frequency (main frequency) of the CPU includes two parts: the FSB and the frequency multiplier. The product of the two is the main frequency. The so-called external frequency refers to the system bus frequency. At present, most of the external frequencies of mainstream CPUs are 66MHz and 100MHz, and AMD's K7 has used external frequencies as high as 200MHz. The full name of frequency multiplication is frequency multiplication coefficient. There is a ratio relationship between the main frequency of the CPU and the external frequency. This ratio is the frequency multiplication coefficient, referred to as frequency multiplication. The frequency multiplier can be from 1.5 to 8.0 in intervals of 0.5. The multiplication of the FSB and the multiplier is the main frequency, so any increase in either can increase the main frequency of the CPU. ?
Earlier overclocking was achieved by changing the multiplier. Generally speaking, increasing the multiplier by 0 to 5 will not affect the stability of the entire machine. Appropriately increasing the frequency multiplier can greatly increase the main frequency of the CPU, which is equal to several levels higher, but such overclocking is not enough to help improve performance. ?
Super FSB is another option with more obvious effect. Super FSB is to increase the system bus frequency, so it is not only a simple increase in the CPU frequency, but also a qualitative leap in the performance of the entire machine. It is precisely because super FSB is more useful than super frequency multiplier, so as long as the combination of FSB and multiplier is reasonably selected, additional benefits can be added without increasing the CPU frequency.
When it comes to external frequency, we have to mention the PCI bus operating frequency. At present, many components such as hard drives, PCI graphics cards, and PCI sound cards on computers use the PCI bus and work at the standard operating frequency of 33MHz. The PCI bus frequency is not fixed, but depends on the system bus speed, which is the FSB! When the FSB is 66MHz, the motherboard uses two-division technology to keep the PCI device operating at 33MHz. When the FSB is increased to At 100MHz, the three-way technology can also keep the operating frequency of PCI equipment from exceeding the standard.
?
But if the external frequency does not use the two standard frequencies of 66MHz and 100MHz, but is fixed at 75MHz or even below 83MHz, the PCI bus can still only use the divide-by-two technology, which makes The operating frequency of the PCI system is 37.5MHz or even 41.5MHz. As a result, many components must work at non-rated frequencies, and whether they can function properly depends on the quality of the product itself. At this time, whether the hard disk can hold up is the most critical, because after the PCI bus frequency is increased, the data exchange speed between the hard disk and the CPU is increased, which is very likely to cause abnormal reading and writing, resulting in a crash (crash). On the other hand, if all devices are fine, then a higher PCI bus frequency can significantly improve the system's operating speed. This is one of the reasons why many overclocking enthusiasts have their own preferences for non-standard FSBs.
The non-standard FSB phenomenon also occurs in FSB systems above 100MHz, but the PCI bus at this time operates at one-third of the frequency of the FSB. Since the overclocking phenomenon is very common at present, although Intel only announced that it will release a 133 MHz FSB CPU, many motherboard manufacturers have begun to express support for 133 MHz FSB and use four-point frequency in systems above 133 MHz. frequency technology to ensure that the PCI bus frequency will not be too high during operation, thus affecting the possibility of final overclocking success. Overclocking the PCI bus frequency is a very serious matter. For detailed information, please refer to the "Overclocking Hard Drive" section in the book.
From super multipliers to super FSBs and then to super PCI bus frequencies, overclocking enthusiasts have moved from the very beginning to a higher level. However, the above-mentioned overclocking methods all depend entirely on the characteristics of the CPU itself. Whether it can be overclocked or not depends solely on the safety margin of the CPU. Such an approach is just to restore the working frequency of the CPU. How can enthusiasts Are you satisfied?
The real leap occurred at this time. Someone actually discovered that the operating voltage of the CPU has a great impact on stability. The operation of the CPU is based on the flow of electrons, and whether electrons can move directionally and stably is determined by the potential difference between the two ends. In other words, a high enough voltage is the prerequisite to ensure the movement of electrons. This is just a not-so-profound theory in electronics, but few people think of it when overclocking. But finally one (several) "genius children" had an inspiration, became interested in the CPU operating voltage, and carried out practical activities with bold and brave ideas combined with careful methods. After countless attempts and choices (some Is this happening?), and finally came to a very, very, very important conclusion: increasing the voltage will help overclocking succeed. Although this is a seemingly simple conclusion, its practical significance cannot be underestimated. In fact, in many cases, by increasing the voltage, the upper limit of overclocking of the CPU can be increased, and the stability is also satisfactory. Changing the CPU operating voltage is a major progress in the history of overclocking. It shows that overclockers have begun to try to get rid of the huge limitations brought by the CPU itself and seek various means to realize the overclocking dream. For most CPUs, as long as the frequency is not restricted by the production process, the CPU can usually continue to increase the main frequency based on the original overclocking by adding voltage. However, the overclocking method of adding voltage is risky, and ordinary enthusiasts are not encouraged to try it. Even overclocking enthusiasts should not adjust the voltage too high (adding voltage is not a panacea, and when the voltage reaches a certain level, Adding more will have no effect).
Whether it is increasing the main frequency or voltage, the heat generated by the CPU will be greatly increased, resulting in electron migration, causing irreparable damage to the CPU. Therefore, as an excellent "overclocking worker", you must pay attention to the heat dissipation of the CPU. Better motherboards now have temperature monitoring devices that can provide CPU surface temperature information. However, the temperature probe may cause inaccurate readings due to various reasons such as improper installation. Therefore, you must not trust the temperature monitoring of the motherboard and relax the heat dissipation work: a real overclocker will never take heat dissipation lightly. Good heat dissipation is essential for overclocking. It’s beneficial and harmless, and there’s no need to worry about it at all. Currently commonly used cooling tools include fans, thermal silica gel, water coolers, etc., which are introduced in detail in the book.
A lot has been said about CPU overclocking. Next, we need to add some information on how to achieve overclocking. Although overclocking is the frequency of the CPU, the CPU itself does not have the function of adjusting the frequency, so all this is achieved through the bridge of the motherboard. Motherboards from the first two years have jumpers or switches to control various optional situations to make the motherboard more compatible and coordinated with other accessories. Jumpers such as FSB, frequency multiplier, and voltage are generally described on the motherboard, and more detailed notes can be found in the motherboard's user manual. It should be noted that not all motherboards will provide all the information about jumpers, and many options need to be sorted and combined to find out. The reason for this situation is that the motherboard manufacturers have the following considerations:
1. The technology is not yet mature and has not yet reached the time of promotion;?
2. There are Risk, not recommended;?
3. Due to external pressure, the logo cannot be made public.
Of course, this is not difficult for DIYers. After a limited number of tests or referring to relevant information, they can discover some "new" FSBs or voltages. Motherboards that use jumpers or DIP switches are inconvenient to use and require certain hardware knowledge, so the more popular approach now is to use soft jumper (no jumper) technology. The soft jumper motherboard can reset the FSB, multiplier and voltage through CMOS, saving a lot of trouble. And all soft jumper motherboards have safety recovery devices that can start normally when overclocking fails. The emergence of soft jumper technology not only greatly facilitates those who install computers, but also makes overclocking more accessible to ordinary computer users, making overclocking no longer the preserve of a few. ?
As mentioned at the beginning of this article, overclocking is not a problem for the CPU family. Overclocking also occurs in other computer components. Generally speaking, overclocking of most hardware is just a passive method chosen to match the CPU. Only graphics cards can have real overclocking. Graphics cards are similar to CPUs in that they both improve performance by changing the clock frequency. However, graphics card overclocking is all achieved through "soft" means, including modifying the registry, using special software, adding statements in Autoexec.bat, etc. , these methods will be specially introduced and will not be detailed here. When overclocking a graphics card, you should also pay attention to heat dissipation issues, and there are certain restrictions. You can't overclock casually (in fact, you can't overclock it). Graphics card overclocking is currently mainly focused on 3D accelerator cards, so it only helps process three-dimensional images and cannot comprehensively improve overall machine performance like overclocking the CPU. ?
Overclocking can greatly improve product performance, allowing users to have products with high cost performance. Especially for CPUs, a small increase in frequency sometimes means saving a lot of money. This is of course something manufacturers are not happy to see. Therefore, Intel, the leader in the industry, uses frequency locking to combat overclocking. The so-called frequency locking is to fix the frequency of the CPU so that the CPU can only work at the rated frequency preset at the factory. At present, Intel has only locked the CPU multiplier. As for the FSB, it has not been touched yet, but I heard that they will lock both of them soon. AMD, another major CPU manufacturer, has never considered frequency locking, leaving everyone with a way out. Once all CPUs are really frequency locked, many friends will lose the fun of overclocking. Only those ultimate players who can manipulate the CPU pins and even produce their own overclockers can continue this sacred practice. Great job. The development of overclocking has become quite popular today, especially in Asia, which has become a very popular thing. There are currently two major overclocking centers in the world: Taiwan and Japan. ?
Taiwan’s reputation in overclocking is very resounding, especially for Chinese people who use the same language. The large amount of overclocking information from Taiwan makes us deeply convinced of the strength of Taiwanese people in overclocking. shift. Since Taiwan is the world's largest production base for boards and peripherals, overclockers in Taiwan have a unique and superior environment. Coupled with the ingenuity of the descendants of Yan and Huang, Taiwanese have made great achievements in overclocking. Taiwan's famous "Overclocker's Paradise" is a website dedicated to discussing overclocking issues. It contains a lot of information and has become a temple in the minds of overclocking enthusiasts. The earliest overclocking trend in China originated from the "Overclocker's Paradise" The reprint of the article shows how influential Taiwan is in overclocking.
?
In contrast, there are very few reports on Japan’s overclocking behavior in China, so the vast majority of Chinese people do not understand Japan’s overclocking situation. In fact, the Japanese are also very interested in overclocking. Although they don't take buying computers seriously based on their income, many Japanese people regard overclocking as a hobby, so in many aspects they are even better than those in Taiwan. A careful comparison of the overclocking situation in Taiwan and Japan reveals that both places have their own preferences. Taiwanese people pay more attention to the matching of accessories and the heat dissipation of the CPU, which means that they should try their best to create good conditions for overclocking and eliminate the negative impact of external factors on overclocking. Many overclockers in Taiwan are amateurs, so their overclocking activities are mainly low-tech. Japanese overclockers are basically semi-professionals, and some are even experts in electronics, so the Japanese tend to focus on hands-on modifications. In particular, the Japanese like to personally make specialized overclocking tools to directly control the CPU frequency. It sounds incredible that the Japanese are purely producing products, but in fact if you are like me and have seen their finished products online, you can appreciate how powerful overclocking is backed by technology and money. Exaggerated and avant-garde.