7 CPU myths that people still believe

Many myths have always been associated with computer hardware. Some of them really make sense in some cases, but there are enough ingrained ones. For example “the heavier the power supply, the better it is,” or “the more video memory, the faster the video card”. And in this article, I will analyze the main myths related to processors.

The higher the frequency, the faster the processor

The myth goes back to the 90s, when many users, so as not to understand the incomprehensible Intel 386, 486 and Pentium, simply looked at the frequency. If some processor had it higher, then it really turned out to be faster. However, now this is generally not true. Processors can have different architectures with completely different performances per hertz. So some Apple A7 with a frequency of 1.3 GHz is at the level of Snapdragon 800 with a frequency of 2.2 GHz and this is not strange. But if we are talking about processors of one generation and one line, then this generally works. For example, an i5-8400 with a frequency of 2.8 GHz is really slower than an i5-8500 with a frequency of 3 GHz.

From overclocking, processors burn out.

It is necessary to distinguish between software and hardware parameters of the processor. So, the frequency is a pure software parameter. For example, for energy saving, it can be reduced to hundreds of megahertz, and with a heavy load, it can fly up to several gigahertz. Therefore, a banal increase in the frequency can not do any harm. At most, you will get an unstable processor operation, but you won’t be able to burn it in such a way.

Another thing is tension. This is a hardware parameter. On the one hand, the higher the voltage, the higher frequencies become available to the processor. On the other hand, each processor has a safe voltage range, and when you exit it there is a non-zero chance to secure a trip to the store for a new CPU.

High temperatures quickly kill the processor

It is believed that working at temperatures close to maximum, the processor will live less. From a physical point of view, this makes sense – at high temperatures, the degradation of a silicon crystal is faster. But there are two important points. First, the critical temperatures indicated by manufacturers are often taken with a good margin of a couple of tens of degrees. Secondly, the life of a silicon crystal is many decades (now there are enough planes from the beginning of the 90s, the “brain” of which are Intel 386 of the same years, and they work fine). So you will not notice a slight decrease in the lifetime during heating having changed the processor much earlier.

But what really can make the processor degrade faster is increasing the voltage to close to critical. In this case, negative effects can be seen after a year. The processor will not be able to work normally at the frequency with which there were no problems when buying, and you will have to reduce it.

The more cores the processor has, the better

It would seem that this is logical. More cores – means more and performance. In practice, it all depends on the specific task. For example, games still can’t really work with more than 8-12 threads. And it may turn out that the top-end 32-core Theadripper will show better performance if you disable half of the cores. So you need to choose a processor not by the number of cores. But by the capabilities of the programs in which you work. Another example is Photoshop, in which a couple of fast cores still produce a much better result than a dozen slow ones. Moreover, there is still enough software that negatively reacts to hyperthreading. When logical cores are turned off, performance may not fall, but, on the contrary, grow.

100% CPU load kills it faster

Not the most common myth – an analogy is usually drawn with a technique that, when working at maximum, wears out and breaks faster. But there are no mechanical parts in the processor. And degradation under normal operating conditions is an extremely slow process, and you will buy a new PC much earlier.

Water cooling of the processor is better than air

From the point of view of physics, everything is true. Water (or most liquids) is a much better conductor of heat than air. However, it should be understood that there are many so-called super-coolers on the market that can divert both 200 and 250 watts from the processor. This is enough for 99% of PC users, and they often cost less than SVO with the same capabilities.

So it makes sense to take a CBO only in two cases. Either you have a powerful processor in a compact package and super-coolers do not fit in it. Or you have a top-end 32-core AMD Threadripper that is clocked at 4.5 GHz and consumes 400+ watts. In all other cases, dropsy usually becomes a waste of money and possible problems in the future.

The specifications on the manufacturer’s website are the last resort

It should be understood that a lot on the manufacturer’s website is written with marketing elements. Of course, there will be no frank lies, but there can be many shortcomings. For example, for the new i9-9900K, a heat package of 95 W is indicated, but in practice, even without acceleration at the maximum TurboBoost frequency, it can consume … as much as 200 W, that is, twice as much. It would seem that Intel is lying? Not at all – with native 3.6 GHz, the processor really fits into 95 watts, and TurboBoost is an optional feature. Therefore, it is better to look at real performance and heat dissipation in reviews.

Leave a Comment