Weak current College】 【overclocked some knowledge-Power By】 【China power house network.

<br> I. What is overclocking? <br> <br> Overclocking is making all kinds of computer components running in higher than the rated speed of the method. For example, if you purchase a Pentium43.2GHz processor, and want it to .run faster, that can be overclocked processors to let it run in 3.6GHz. <br> Solemn declaration! <br> Warning: Overclocking may cause the part to scrap. Overclocking risks if overclocked it whole computer's life may be shortened. If you try .overclocking, I will not use this guide and be responsible for any damage caused. This guide is only for those generally accepted this overclocking Guide / FAQ and overclocking of possible consequences. <br> Why want overclocking? Yes, the most obvious motivation is to .get from the processor than paying more in return. You can purchase a relatively inexpensive processors and overclocking to run in the more expensive the speed of the processor. If you are willing to invest time and effort, Overclocking can save a lot of money; if .you are a fanatic like me, overclocking players can bring you more than likely buy from stores to the faster processor. <br> <br> Second, overclocked dangerous <br> <br> First of all I have to say, if you are .careful and know what to do, to you, to your computer through the overclocked to cause any permanent damage is very difficult. If the system-too, burned down or fail to start. But just give it to the limit is very difficult to burn .-out system is still however the danger. First and most common risk is fever. In the computer part higher than rated parameters runs, it will produce more heat. If you do not have adequate cooling, the system might overheat. However General overheating is .not able to destroy the computer. Heat only the computer scrap is repeatedly attempt where the computer is running in a temperature higher than recommended. As I said, you should try to suppress in 60C following. <br> But without excessive worry about overheating problems .. In the event of a system crash will have symptoms before. Random reboot is the most common symptoms. Overheating is easy through the use of thermal sensors to prevent, it can display the temperature of the system is running. If you see the temperature is .too high, or at lower speeds run system or better heat dissipation. Later I will be discussed in this guide. <br> Overclocked another "dangerous" is it possible to reduce part of life. In part, impose a higher voltage, it's life .will be reduced. Small upgrade will not cause too much, but if you intend to make substantial overclocking, it should be noted that the shortened life expectancy. However this is usually not a problem, because any overclocking are less likely to use the same parts .as four or five years, and it is not possible to say anything as long as the compression will not be able to stay on 4-5 years. Most processors are designed for maximum use of 10 years, so in overclocked to mind the loss of .some years, in return for performance of increase are usually worth it. <br> Basics <br> <br> In order to understand how to overclock your system, you must first understand how the system works. Used to overclocking the most common part .is the processor. <br> <br> In the purchase of processor, or CPU time, you will see it run faster. For example, to run under the 3200MHz Pentium43.2GHzCPU. This is a second processor has undergone a number of clock .cycles of the measure. A clock cycle is a period of time, during which time the processor to execute a given number of directives. So logically, the processor in a seconds to complete the more clock cycles, it can be faster to process information, .and the system will run faster. 1MHz per second one million clock cycles, so the processor in every 3.2GHz seconds can experience 3,200,000,000 or 3 billion 200 million clock cycles. Pretty amazing, isn't it? <br> <br> Overclocked aims .to improve the level of GHz processor so that it can experience per second more clock cycles. The formula for calculating the processor speed is this: FSB (in MHz) x multiplier = speed (in MHz). Now to explain the FSB and multiplier is .what: <br> <br> FSB (on AMD processors is HTT), or the front side bus, is the entire system and CPU communication channel. Therefore, the FSB will run faster, it is clear that the entire system can run faster .. <br> <br> CPU manufacturers have found increased CPU FSB effective rate method. They just in each clock cycle sends more instructions. So the CPU manufacturers already have each clock cycle sends two directives (AMDCPU), or even each clock cycle four .directives (IntelCPU), rather than sending each clock cycle one instruction. Then consider the CPU FSB speed reading, you must realize that it's not really at that speed. <br> <br> IntelCPU is "four-core", that is, .they send each clock cycle 4 of the directive. This means that if you see the FSB 800MHz, potential FSB speed only 200MHz, but it sends each clock cycle 4 instruction, reached the effective speed 800MHz. The same logic applies to AMDCPU, however they .are just "core", meaning they each clock cycle only send 2 instruction. So in the FSB on 400MHz AMDCPU by potential 200MHzFSB per clock cycle to send 2 instruction. <br> <br> This is important because when the overclocked CPU to handle .real FSB speed, but is not a valid CPU speed. <br> <br> Speed of multiplier section which is a number, multiply by FSB speed gives the overall speed of the processor. For example, if you have a heart with 200MHzFSB ( .in x 2 or x 4 the real FSB speed before) and 10 times the frequency of CPU, then the equation becomes: (FSB) 200MHz × (multiplier) 10 = 2000MHzCPU speed or 2.0GHz. <br> <br> In some CPUs ., such as Intel since 1998 of processors, doubling the lock can not be changed. In some, such as AMDAthlon64 processor, frequency capping is "locked", that is, you can change the multiplier to a lower number, but not to much higher .than the original. On the other, of the CPU multiplier is completely let go of, means being able to change it to any desired number. This type of CPU is Overclocked Acura, because you can simply by improving the multiplier to overclocked CPU, but .now very rare. In the CPU to increase or decrease the multiplier is much easier than the FSB. This is because the multiplier and FSB, it only affects the CPU speed. Change the FSB, you are actually changing each individual PC components to communicate with .the CPU speed. This is overclocked systems all other parts. This is not going to be super overclocked parts too high and could not work, you may bring a wide variety of issues. But once you understand how the overclocking, you'll know how .to prevent these problems. <br> <br> In the terminology of FSB AMDAthlon64CPU is really a misnomer. Essentially no FSB. FSB is integrated into the chip. This allows communication with the CPU FSB than standard Intel FSB method is much faster. It .also may cause some confusion, because the Athlon64 on FSB may sometimes be described as HTT. If you see some people are talking about raising the HTT Athlon64CPU and is recognized as a normal discussion FSB speed speed, then the HTT as FSB to consider. For .the most part, they run in the same way and can be regarded as the same thing, but to be considered as FSB HTT to eliminate some of the confusion may occur. <br> <br> III. how to overclock <br> < .br> You now know how to get the processor's rated speed. Very good, but how to improve the speed? <br> <br> Overclocking the most common way is through the BIOS. At system startup by pressing a specific key to enter the .BIOS. Used to enter the BIOS of the most common key is the delete key, but some might use as F1, F2, F button, Enter and what key. The system begins to load Windows (any use of the OS), you should .have a screen at the bottom of what key you want to use. <br> <br> Overclocking BIOS support is assumed, that once into the BIOS, should be able to use the overclocked system requires all settings. Most likely to be adjusted settings .are: <br> <br> Multiplier, FSB, RAM, RAM speed and delay RAM ratio. <br> <br> In the most basic level, you only have to do is get you to the highest FSB × multiplier formula. Complete .the easiest way is to increase the multiplier, but that in the majority of the processor is not possible, because the double frequency is locked. The second way is to improve the FSB. This is quite a limitations, all while improving FSB RAM issues must .be addressed are described next. Once you have found the CPU's speed limit, and there was not only a choice. <br> <br> If you really want to put the system in pushing the envelope, in order to put the FSB rises higher .frequency can be reduced. To understand this, imagine having a 2.0GHz processor, it uses 200MHzFSB and 10 times the frequency. Then 200MHz × 10 = 2.0GHz. Obviously this equation works, but there are other ways to obtain 2.0GHz. You can .raise to the multiplier and FSB 20 drops 100MHz, or can put up to FSB 250MHz and doubling down to 8. These two combinations will provide the same 2.0GHz. It is not two combinations should provide the same performance? <br> <br> .This is not the case. Because the FSB is the system used to communicate with the processor, the channel should make it possible. So if you put the FSB to 100MHz and multipliers up to 20 words, still owns 2.0GHz clock speed, but the .rest of the system and processor communication will be much slower than before, resulting in the loss of system performance. <br> <br> Ideally, in order to improve the FSB as possible heights should decrease the multiplier. In principle, this sounds very .simple, but in other parts of the system will become more complex, because the rest of the system is also determined by the FSB, first and foremost is the RAM. This is what I want in the next section. <br> <br> .Most retail computer manufacturers use does not support the overclocking motherboards and BIOS. You will not be able to access from the BIOS settings required. There are tools that allow overclocked from Windows system, but I do not recommend using them, because I've never .personally tested. <br> RAM and its effect on overclocked <br> <br> As I said before, the FSB is system and CPU communication path. So improving the FSB also effectively overclocked systems of rest parts. Affected by the increasing influence of .the FSB is the largest part is RAM. Buy RAM, it is to be set at a certain speed. I will use a table to display the speed: <br> <br>. PC-2100-DDR266 <br>. PC- .2700-DDR333 <br>. PC-3200-DDR400 <br>. PC-3500-DDR434 <br>. PC-3700-DDR464 <br>. PC-4000-DDR500 <br>. PC-4200-DDR525 <br .>. PC-4400-DDR550 <br>. PC-4800-DDR600BR> <br> To understand this, you must first understand how the work of the RAM. RAM (RandomAccessMemory, random access memory) is used as the CPU needs fast .access to files in temporary storage. For example, when loading the game plane, the plane load the CPU to RAM so that it can in any need quick access to information, rather than from a relative slow hard disk load information. <br> < .br> The important thing to know is the RAM to run at a certain speed, the ratio is much lower CPU speeds. Today, most RAM to run in 133MHz 300MHz between speeds. This may be confusing, because the speed is not listed in my .diagram. <br> <br> This is because RAM manufacturers follow the CPU manufacturers approach, trying to RAM in each RAM clock cycle information sent twice. This is the list in the DDR RAM speed rating. It represents the DoubleDataRate (double data rate .). DDR400 RAM therefore means that the effective speed 400MHz operate, DDR400 in 400 represents the clock speed. Because it sends each clock cycle two directives, it means that it is really working frequency 200MHz. It's like AMD's "core" of the FSB. .<br> <br> So back to RAM. There are lists of prior DDRPC-4000. DDR500 PC-4000 equivalent to, that means PC-4000 RAM speed with effective 500MHz and potential 250MHz clock speed. As I said before, in improving .the FSB when Overclocking a system effectively in all things. This also includes RAM. Rated at PC-3200 (DDR400) of RAM is run at the highest speed of 200MHz. For not overclocked, this is enough, because the FSB in any case not .more than 200MHz. <br> <br> But when you want to put the FSB rose to more than 200MHz speed when problems arise. Because RAM only rated to run at maximum speed, 200MHz increases FSB to above 200MHz may cause system crash. This .how to resolve?, there are three solutions: use FSB: RAM, RAM or purchase rated overclocked at a higher speed of RAM. <br> <br> Because you may just learn the three choices in the last one, so I will explain .them: <br> <br> FSB: RAM ratio: If you want to raise to the FSB support than the RAM of high speed, you can choose to have the RAM runs at lower than FSB speeds. This use of FSB: RAM ratio .. Basically, FSB: RAM ratio allows you to select numbers to FSB and RAM speed ratio between the establishment of a. Assuming you are using is PC-3200 (DDR400) RAM, I mentioned it before running the 200MHz. But you want to improve .the FSB Overclocking CPU to 250MHz to. Obviously, the RAM will not support elevated FSB speeds and is likely to cause system crash. To solve this, you can set up a 5: 4 ratio of FSB: RAM. Basically this means that if you .run under the 5MHz FSB, RAM will run only under the 4MHz. <br> <br> More simply, the 5: 4 ratio as the ratio of 100: 80. Then run for the FSB, the 100MHz RAM will run only under the .80MHz. Basically this means that the RAM will only run in the FSB speeds of 80 per cent. So as far as the objectives of the FSB 250MHz, run in 5: 4 ratio of FSB: RAM, the RAM will run at 200MHz, that .is 80% 250MHz. This is perfect, because the RAM was rated at 200MHz. <br> <br> However, this solution is not ideal. To run a ratio of FSB and RAM leads to FSB and the time difference between RAM communication. .This caused a slowdown, and if the RAM and FSB running on the same speed, then do not occur. If you want to get the system's maximum speed, use the FSB: RAM ratio will not be the best option. <br> <br .> 4. voltage and how does it affect the overclocking: <br> <br> In a pole when overclocking, regardless of what to do or have more good heat will no longer increase the speed of the CPU. This is most likely because the .CPU does not have sufficient voltage. As previously mentioned the memory voltage is very similar to the situation. To solve this problem, as long as the increased CPU-voltage vcore, that is. To RAM the section describing the same way to accomplish this. .Once you have made sufficient voltage CPU stability, you can either let the CPU in the speed or try to further overclocked it. As with the handling of RAM, be careful not to overload the CPU voltage. Each processor has manufacturers recommended voltage settings. Find .them on the Web site. Try not to exceed the recommended voltage. <br> <br> Remember to improve CPU voltage will cause much more heat. This is why in overclocking, good heat dissipation of nature. The boot from the next topic. .<br> <br> Cooling: <br> <br> As I said before, when the CPU voltage, calorific value growth. This will require appropriate cooling. Basically there are three "levels" of chassis cooling: air cooled (fan) ., water cooled, Peltier / phase change cooling (very expensive and high-end thermal). <br> <br> I on the Peltier / phase change cooling method is really not much to know, so I'm not going to say it .. You only need to know is that it will spend more than $ 1000, and enables CPU maintained at a temperature below zero. It is intended for very high-end overclocking to use, I would like to no one will use it. However, .in additionBoth cheaper and more practical. <br> <br> Everybody knows that the forced-air cooling. If you are in front of the computer, you may hear from its outgoing continuous hum. If you look at it from behind, you will .see a fan. This fan is basically all the wind cold: use fans to draw cool air and exhaust the hot air. There are various methods for installing the fan, but usually there should be equal to the amount of air suction and discharge. Water .cooling more expensive than the air cooling and singular. It is essentially using pumps and tanks to system cooling, more effective than that of air cooling. <br> <br> Those are the two most commonly used chassis thermal methods. However, good chassis .cooling on a cool computer, is not the only required parts. Other major components are CPU heatsink / fan, or HSF. HSF's objective is to heat from the CPU and to enter the chassis to boot, so that it can be discharged from the chassis .fan. The CPU has been an HSF is necessary. If you have a few seconds without it, the CPU may be burnt down. <br> <br> 5. If the computer does not show up (post no show), what should .I do? <br> <br> Depending on your motherboard. "Failure recovery" option is used to reset the CMOS, usually by jumper discharge. In the motherboard manual to find details. If you are overclocking too high but BIOS settings intact, .most recently in a fever-motherboard have a option to reduce the frequency of display, so you can enter the BIOS and lowered to the stable operation of clock speed. <br> <br> On some motherboards, the computer by opening the Insert key .when you press and hold to finish (normally must be a PS / 2 keyboard). If the computer passes through efforts is not yet before the show, some will automatically reduce the frequency. Sometimes computer does cold start (press the power button shows) .but after a while the keep will run, then restart. On other occasions, the computer will be nicely cold start, but not warm start (restart). Those are signs of instability, but if you are satisfied with the stability and ability to deal .with the problem, then it usually does not cause big problems. <br> <br> 6. what limits overclocking? <br> <br> Usually RAM and CPU is the only significant constraints, particularly in the AMD system due to memory run .asynchronously and inherent problems (see the following sections of the FSB). RAM had to run at the same speed with FSB or its crossover frequencies. Memory can be run at higher speeds than the FSB, rather than just below it. However there is a .running higher delay / higher memory voltage selection, it is becoming more and more like a limiting factor, particularly as the new platform (P4 and A64) from the asynchronous operation has less loss of performance. <br> <br> CPU has become a .major limiting factor. Unique processing cannot run faster CPU is to increase voltage, but more than the maximum core voltage will shorten the life of the chip (although the overclocking will do), but sufficient thermal part of the solution to the problem. <br .> <br> With the use of too high core voltage of another problem in the P4 platform to SNDS, or SuddenNorthwoodDeathSyndrome (sudden death syndrome), use any of the above 1.7v causes the processor voltage rapidly and premature end-of-life, .even if the use of phase change cooling. However, the new C-core chip, that is, the EE chip, and Prescott chips do not have this problem, at least in the range. The heat can also impede the overclocking, too high .a temperature will cause instability. But if the system is stable, then the temperature is usually not too high. <br> <br> 7. now already overclocked many, what to do? <br> <br> If you want to run .some benchmark testing. Let Prime95 (or you choose to emphasize testing-complete depending on your) running the full length of time (usually 24 hours no failure is that the system is stable). <br> <br> 8. What is the .FSB? <br> <br> FSB (front side bus or FrontSideBus,) is the easiest and most overclocked common methods. FSB CPU and system is the rest of the connection speed. It also affects the memory clock, it is the speed of .the memory to run. In General, the FSB and memory clock, the higher is equal to two, the better. However, in some cases this is not established. For example, let the memory clock run faster than the FSB will not have any .real help. Similarly, in the AthlonXP system, FSB runs at a higher speed and force the memory and FSB is not synchronized (use will discuss memory divider) on the performance of running hindering than at a lower FSB and synchronous memory to be much more .serious. <br> <br> FSB in Athlon and P4 systems involved in different ways. In this way, it is the Athlon DDR bus, means that if the actual clock is 200MHz, it's running in 400MHz. In P4, which is the ."core", so if the actual clock is the same 200MHz 800MHz, representative. This is Intel's market strategy, because the average user, the higher is better. Intel's "four-core" FSB effectively have a real advantage to small performance loss expense .allows P4 chipset and memory are not synchronized correctly. Higher per clock cycle speed makes it more opportunities for memory cycle and CPU cycles coincide, the equivalent, the better the performance. <br> <br> 9. Why make PCI / AGP bus run .super specification can cause instability? <br> <br> Let the PCI bus run super spec lead to instability is mainly because it enforces a very strict allowed deviation of parts run on different frequencies. PCI specifications are usually require 33MHz. Sometimes it provides 33.3 .MHz, I believe that is close to trueAs specifications. High speed PCI are the main victims of the hard disk controller. Some controller card has more than any other card higher permissible deviation, you can run the increased speeds and no obvious damage. <br .> <br> However, in the majority of the motherboard's onboard Controller (especially the SATA controller) for high speed PCI is extremely sensitive, if the PCI bus operating at 35MHz down there is damage and data loss. Most able to address, in fact ., Super specifications 34MHz ranges less than 1MHz (depending on board how rounded to 34MHz ... For example, most boards may 134 to any 137MHz FSB report under 34MHz PCI speeds. Actual range from 33.5MHz to 34.25MHz and may be based on the .motherboard clock frequency change and variation. In higher FSB and higher dividers, the range might be larger). <br> <br> Sound cards and other integrated peripherals in PCI bus runs Super specifications also suffer. ATI video card on AGP nVidia card is .faster than the much smaller tolerance (direct link to the PCI speeds). Remember, most RealtekLAN card (based on the PCI expansion slot, and consume) is set in from 30 to 40MHz any frequencies between safe operation. <br> <br> .10. What is a frequency-doubled? <br> <br> Harmonic combination of FSB to determine chip clock speed. For example, the double frequency collocations 12 200 of FSB will provide 2400MHz clock speed. Like in the previous overclocked chapters, some .CPU is lock multiplier and some not, that only certain CPU frequency regulation allowed. If you have a multiplier adjustment can be used either in the FSB restricted motherboard for higher clock speed, or when the chip is restricted for higher FSB. <br> < .br> 11. What is the memory frequency? <br> <br> Memory frequency to determine the memory clock speed ratio of the FSB. 2: 1 of FSB: RAM frequency will receive 100 RAM clock on the FSB 200MHz. Frequency of the .most common use is to run in the P4C systems 250FSB PC3200RAM, 5: 4 ratio. In most Intel systems have 4: 3-n and 3: 2 ratio. Athlon system using frequency is not like the P4 systems so efficient use of memory, .as part of the description above FSB. Memory frequency should be used only to obtain stability, instead of 1, because even in the P4 it also undermines performance. If the system does not have to take memory frequency is stable (or if the memory voltage .increase can solve the problem), that do not use a frequency. <br> <br> 12 different memory latency mean? <br> <br> CAS latency, sometimes also known as CL or CAS, the RAM must wait until it can .once again read or write the number of minimum clock. Obviously, this figure, the lower the better. TRCD is memory special line of data is read / write delay before. This figure is lower, the better. <br> <br> TRP .is primarily the time line charge. TRP is written to the row in the data after another row is activated the wait time before. The lower the better. TRAS is row is activated in minimum time. So basically tRAS refers to the number of rows must .be open within time. This figure with the RAM settings, change quite a lot. <br> <br> 13 different memory levels mean? (PC2100/PC2700/PC3200 etc.) <br> <br> Level directly refers to get maximum .bandwidth, and indirect means memory clock speed. For example, PC2100 owns 2.1GB / s maximum transfer rate, and 133MHz clock speed. As another example of PC4000, with 4 GB / s ideal transmission speed and 250MHz clock. To get from PCXXXX grade ., the clock speed rating is divided by 16. The speed rating took 16 got bandwidth level. <br> <br> 14. how to express the actual DDRXXX memory clock speed? <br> <br> DDRXXX is just the actual clock speed .; that is, is set in 200MHz DDR400. If you want to know DDRXXX speed PC-XXXX speed, it took 8. <br> <br>.