Computer

Underclock why is it used on Intel and AMD CPUs or NVIDIA GPUs?

We are living in a time where both GPUs and CPUs are scaling towards two totally different aspects, necessary and that on the other hand, are not new at all. We are talking of course about a greater number of cores / Shaders and a higher frequency, but this is causing consumptions and temperatures to skyrocket to levels above what many are willing to cool. Here comes a technique as old as the very existence of the system clock: Underclock, but what is it?

Old ways never die and as long as they are available many will take advantage of them. For various reasons there are users who pretend the opposite of what Intel, NVIDIA or AMD achieve and this is precisely where everything becomes interesting, because it is a struggle to achieve the best performance with a series of requirements.

Underclock, as old as necessary for CPUs and GPUs

Processor clocking switching speed

We are looking for the highest performance and for this everything is based on few factors, such as frequency, IPC or number of cores in the die, but what if it turns out that we need one of them but not the frequency itself?

Then it only remains to underclock, which is a very simple technique that allows CPUs and GPUs to lower the clock speed regardless of factors such as performance. Logically, if we want to lower the frequency, we will lose performance in our processor or graphics card, just like in the memories, but there are plenty of reasons to achieve it at the same level that any user would undervolt these components.

Underclock vs undervolt, the same in essence? Do they serve the same purpose?

If we compare these two techniques (underclock vs undervolt) the differences are simple and clear: with the first we will dedicate ourselves to lowering frequencies thereby losing performance, with the second we lower voltages thereby losing stability and possibly if we are very aggressive also performance.

Underclock does not imply undervolt or vice versa, they are two totally independent techniques and yet they can and should be used in unison as long as you know what you are doing. They are techniques that require a medium-high level of knowledge when we get deeper into them.

To understand the concept of underclock better, we could say that it is the opposite of the so famous overclock and obviously its effects live up to its meaning, therefore, as it happens in overclock and its voltage that normally it is necessary to raise it (this does not have a specific name as such, everything is included within overclock) in underclock, it would be necessary to apply undervolt to compensate and gain more general advantages.

Why underclock if what we are looking for is more performance?

Well, although it may not seem like it, the current paradigm has shifted to some rather strange levels. The Mainstream CPU range is getting all the praise and attention from the user, while the HEDT range of processors is being residual and possibly disappearing in the long run, because core and frequency levels are skyrocketing to look like servers from just 5 years ago. .

Therefore, it is such a curious case that users are buying hardware with such extremely brutal performance that they do not think about other issues adjacent to the purchase of a processor or graphics card.

We refer to things that we have all seen as soon as we are within the professional sector: i9 or Ryzen 9 CPUs with 100 euro motherboards, graphics cards with consumptions above 300 watts with PSUs of just 600 watts, boxes with more cooling that deficient because they are low-end, 20 euro heatsinks for TOP processors and similar things.

Some may be throwing their hands to their heads and with good reason, but that is the real world of the average user, who sometimes allows themselves to be advised or poorly advised by people who have little or no knowledge of the sector and of course, when the problems we seek information.

But then it is too late because the money is spent and the problem is very real. The symptoms are clear: PC slowdowns, reboots, freezes, feeling of low speed etc etc … These are mostly due to thermal or electrical problems, where if we have already spent all the budget money and we cannot change certain parts of the PC, the solution goes through the most optimal configuration for our needs.

Lowering the frequency avoids some of the problems we may have

If we cannot refrigerate the component in question there are two options: underclock or undervolt. The problem is that the second requires certain knowledge and a lot of testing time to affirm that there is stability, while the first only needs to touch one or two parameters to guarantee the same stability or more (if there are problems of course).

For this reason, the difficulty makes the average user without knowledge about these issues end up choosing to choose underclock over undervolt, because getting involved in passing tests at each voltage drop and adjustment of settings is not for everyone, but it does modify two parameters of quickly in BIOS.

These parameters to be played for the underclock vary depending on the CPU or GPU in question. For unlocked processors for overclocking it is as simple as changing the multiplier of it, but for blocked processors we will have to touch the formerly called FSB, now called BCLK.

In the first case, only the processor will lose performance, while in the second we will have some added problems, since the PCIe will be affected in its speed on some platforms. The question now is how much do you have to lower the multiplier or the BCLK?

The answer is as much as is necessary to cover with what we need. That is, if we have a temperature problem, we will have to reduce the frequency to levels where we feel comfortable and the processor is safe and its protection does not come into play.

If it is a GPU, it is a little more of the same, except that in this case we can only lower MHz through third-party programs for Windows. The good thing about this technique is that it is simple, the voltage will remain as is and therefore we will not lose general stability, passing us does not give any advantage, but rather the opposite.

In short, we must spend time on what is important, keep the processor in a safe temperature range, the same GPU, at the cost of having a lower performance.

The post Underclock why is it used on Intel and AMD CPUs or NVIDIA GPUs? appeared first on HardZone.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *