Overclocking has always been an attention-getter in the PC enthusiast world. Motherboard manufacturers have taken notice of the growing do-it-yourself-er population and have even released special revisions of their motherboards specifically geared towards overclockers. This accomplishment is very impressive because, when you think about it, the majority of the profits that motherboard manufacturers could potentially make come from OEM deals and not from the PC enthusiast market. Just as motherboard manufacturers have realized, this market is a very fast paced one and by getting involved in it now they are opening their arms to an entirely new population of eager buyers.
AnandTech readers are already probably familiar with overclocking and its origins, but for those of you that aren’t, overclocking is the simple process of running any component of your system at a frequency that is higher than the component’s rated frequency. The most common form of overclocking involves increasing the frequency of your CPU, which is the topic at hand in this article.
Back in the days of the 386, overclocking could only be accomplished with the aid of some intense cooling measures. It wasn’t until the release of the 486 that overclocking became more mainstream, especially with the later releases of the CPU. AMD and Cyrix also had their periods of overclocking friendly time with their 5x86 chips that would plug directly into the socket on most 486 motherboards. For example, the AMD 5x86 running at 133MHz was the basis for many overclocking success stories during its time since increasing the Front Side Bus (FSB) frequency of your setup from 33MHz up to 50MHz and dropping the clock multiplier down to 3.0x would result in a reasonable overclock to 150MHz which was quite impressive at the time.
Fast forwarding to more recent times, the release of Intel’s Celeron changed the face of overclocking, at least with Intel CPUs, for quite some time. While the Celeron, running at 266MHz was often times able to be overclocked to 400MHz (by increasing the FSB frequency from 66MHz to 100MHz) which resulted in a very impressive 50% overclock, the biggest impact it had came from the fact that it was the first Intel CPU to feature a locked clock multiplier. From that point on, every single Intel CPU was manufactured with a locked clock multiplier that, to this date, has yet to be broken in a manner that has been publicized. The reason this affected the overclocking world so incredibly was because this limited the number of options you had for overclocking Intel CPUs Now, you could only rely on increasing the FSB frequency to overclock your CPU which is more difficult than simply increasing the clock multiplier because most of your motherboard runs off a multiplier or divider of your FSB frequency; therefore, increasing the FSB frequency overclocks much more than just your CPU.
Why would Intel be so determined to stop overclocking that they would resort to shipping multiplier locked CPUs? The answer to that trick question is that Intel wasn’t determined to stop overclocking, they were out to stop remarking.
In the grand scheme of things, the 1 – 2% of computer users that actually overclock doesn’t matter at all to Intel. However, the unscrupulous vendors that overclock CPUs and sell them as higher speed CPUs without first informing the customer that they are overclocked is what Intel has zero tolerance for. This process is known as remarking and it grew to a very big problem before the release of the Celeron. The advent of multiplier locking on CPUs took a big chunk of remarkers out of the game In addition, Intel’s strict enforcement of federal laws helped federal officials raid remarking facilities and their utilities that report to end users whether or not their CPU is running at an overclocked speed has made remarking Intel CPUs much more difficult than it once was.
On the other hand, AMD has never shipped a multiplier locked CPU.