It consists of that itself, of raising the clock speed over the face-value of the mike. This practice can be realized by the way or have been a victim of a trick, as; in any case, it contains risks for the mike overclockeado.

I insist: it contains risks for the mike; do not say that it did not warn him itself. The mikes of the same class are born, in general lines, all equal ones. Then they are proved and one classifies them with such and such speed, according to the demand of the market and what it has been proved that they resist without any mistake.

This means that many mikes can be used at more speed than they mark, although out of specifications and therefore of guarantee. The negative consequences are three:

1. that does not work at any more speed of the pronounced one (since not at all, there is left he as it comes and in peace);
2. that gets damaged (rarely it happens if it is raised in a staggered way and watching if it fails);
3. that works but it warms up (it will always happen; on having gone more rapid, it generates more heat).

The computer science is an area paid for the falsifications. Without the serigrafiado all the chips seem equal, and it is impossible to know his mark, model or speed.

This is applied to the perfection for the mikes; formerly it was strange that someone was selling a falsified mike, but with the arrival of the Pentium the problem went so far as to acquire alarming dimensions, with bodies like Interpol mobilized in the hunting of the forger.

The most typical tricks are usually:

• Falsification in itself: a chip is taken, one erases or he covers the serigrafiado and he writes himself above, obtaining a more expensive new chip. At first the falsification was very vulgar, and a little of acetone was revealing it; nowadays, an expert cannot even be sure.
  It is very difficult to prevent, how it is not coming to companies of recognized prestige where the risk is minimal or choosing a cheap chip, which sure that they have not forged. Also can be required that it should be a chip not OEM, but with its own box and sealed guarantee, but these chips are much more expensive (and the box can be false...)
• Mikes exchange: one day occurs to us to raise the fan of the mike and: surprise!, it is a normal Pentium instead of MMX; or it goes at another speed, for example. It goes away to the shop and they say to us that it is an error, which they do not understand what can have happened... and how we do not want trouble and is possible that they are right (in whose case what there are is a few misers), we leave that they change the chip to us and go away.
• Changes of mark: slightly very common in the epoch 386 and 486, now much less. We were paying a mike Intel and they were selling to us an AMD, Cyrix or Texas Instruments; mikes that sometimes are equal or better (me 386 of AMD was fantastic, and it might still work without problems) but that are cheaper, therefore the saving is for the seller, not for us.

Aberrant histories are known, I insure it myself; Pentium MMX slower than 486, mikes with 5 stickers on the serigrafiado... false, clear; and even worse things. The solutions, next: