Learn about the technologies behind the Internet with The TCP/IP Guide!|
NOTE: Using robot software to mass-download the site degrades the server and is prohibited. See here for more.
Find The PC Guide helpful? Please consider a donation to The PC Guide Tip Jar. Visa/MC/Paypal accepted.
|View over 750 of my fine art photos any time for free at DesktopScenes.com!|
Why Is Overclocking Even Possible?
To someone who doesn't understand a great deal (yet) about how PCs work, it may seem strange that you can run parts of the system faster than they are intended to run. Isn't a Pentium with MMX 166 always going to run at 166 MHz? The answer is "no". There are many parts of the PC that have the capability of running at different speeds. This is built into the hardware primarily for flexibility, to allow many different pieces of hardware to work together, and to make motherboard design simpler and cheaper.
The actual speed of your system bus and processor are controlled not by the processor itself, or by the chipset (which manages the system bus) but by jumpers on the motherboard. When you change these jumpers, you change the speed of the system bus or the processor. This allows many different speeds of processors to work on the same motherboard, and also allows chip makers to not have to have different internal designs for some of their chips. Some newer "jumperless" motherboards substitute special BIOS settings for these jumpers, but the concept is identical.
As discussed here, processor manufacturers always try to make their product as fast as possible, because faster hardware sells for more money. However, some chips turn out to be able to run at higher speeds than others. Each chip is tested to see how fast it will run, and the ones that will run faster are given faster labeling. This is called speed rating. So in a given family, some chips coming from the same wafer may become AMD K6-200s, and others AMD K6-166s.
The reason that overclocking works is that reputable companies are conservative in their speed rating. If Intel sends a chip out labeled as a Pentium with MMX 233, it wants to be darn sure that the chip will run at that speed. If there is any doubt of the chip's capabilities, it will be sold instead as a Pentium with MMX 200. Overclockers attempt to exploit this conservatism, sometimes with success and sometimes without. Occasionally a chip labeled at a lower speed will run at a higher speed, and sometimes it won't.
There is also evidence that in many cases, the processor manufacturer will at times intentionally underrate chips in order to meet market demand and create differentiation between high-end and low-end product. When a chip is new, the manufacturing process probably legitimately creates a small yield of faster parts and a high yield of slower parts. As the chip matures, more high-end chips are created. It is possible that if "too many" high-end chips are produced to meet demand, some may be marked at the lower speed to fill orders. Furthermore, sometimes dedicated production lines are used to produce chips of different speeds, but they are produced at rates not matching demand forecasts and may be marked at a lower speed. These chips of course would be easily "overclockable" because they were never marked at their full potential to begin with. The problem, of course, is that it is virtually impossible to figure out which chips were underrated! Still, a chip is generally more overclockable when it is the millionth off the production line than the first.
The latest craze is overclocking the system bus. More accurately, this is overclocking the system chipset and expansion cards. This is made possible because recently, motherboard manufacturers have started putting non-standard bus speed settings on their motherboards, allowing you to exceed the speed rating of the chipset.