Possibilities for the overclocking
Bearing in mind the theory explained in the previous point, we have 3 different possibilities to realize the overclocking:
(1) To rise the multiplier of the mike.
By means of this method we will change only the internal speed of the mike, while the day pupil will remain constant. Thus the only element that he suffers is the mike, while other devices work at his normal speed.
We will have to study the manual of the motherboard and see how the change of the multiplier is realized; in the most habitual case so called jumpers will be done by means of a few small microswitches, although in the most modern basic badges it is realized by software, generally inside the BIOS (in this the badges ABIT were the first ones). Even, at the worst, it is possible that the badge detects automatically the theoretical characteristics of the mike and does not allow us to form them to hand... as it happens in any basic badges made by Intel.
Another problem that happens at present is that almost all the mikes Intel from 300 MHz Pentium II (including Pentium III and Celeron "Mendocino"), as well as some previous series, they have the multiplier limited to a few concrete values or I fix to the only value, for example 5x for a 333 MHz Pentium II (5x66).
In such a case, and if we do not have a mike AMD (that or they take the free multiplier as a defect or it can be liberated, as we will see), we will have to try to:
(2) To raise the speed of the bus.
Doing this we will increase the speed both of the microprocessor and of other elements of the computer (the motherboard, the memory, the cards of expansion...). Again, it will be formed by means of jumpers or in the BIOS.
For example, the "official" configuration of a 400 MHz Pentium II is 4x100; if we put a 110 MHz bus , the overclocking would have the following effects:
Observe how the speed of the different elements is usually related to that of bus (or day pupil) of the mike:
It is a method that can give a lot of game, although again only realizable with mikes without fixed multiplier. We can do authentic marvels, and even manage to accelerate the computer without changing the internal speed of the mike so that he does not suffer, for example changing a Pentium 150 of 2,5x60 to 2x75, what would accelerate the bus PCI and the memory without risk for the mike.