As it turned out after the announcement of the new flagship Intel Core X processors, motherboards of the first wave based on the X299 chipset are poorly suited for Skylake-X overclocking due to the critical heating of the power converter. Fortunately, some motherboard manufacturers did not deny the problem and responded to criticism. So, in the new video, the famous overclocker Der8auer, who first drew attention to overheating, says that ASUS has changed the design of the cooling system of the flagship board ROG Rampage VI Apex.

At the end of June, enthusiasts got the opportunity to get acquainted with those serial motherboards that manufacturers have prepared for the new Core X series processors. However, such an acquaintance could hardly be called pleasant. Unexpectedly for themselves, many overclockers are faced with the fact that the motherboard power converter overheated even with a slight overclocking of the processor. As it turned out, motherboard manufacturers did not pay due attention to improving the heat dissipation from the VRM and applied their usual approaches to its cooling. And this means that beautiful, but inefficient aluminum radiators with a mediocre thermal interface were used everywhere. And if earlier manufacturers got away with it, then with the older 10-core Skylake-X processors, which can consume more than 300 W during overclocking, such cooling, as it turned out, can no longer cope.

ASUS was the first manufacturer to acknowledge the problem. In this regard, the flagship board ROG Rampage VI Apex, which luckily did not have time to reach the stage of mass production, will come to the market with a modified VRM cooling system.

Initial version of ASUS ROG RAMPAGE VI APEX

In the new version of the board, instead of a heatsink made of two solid aluminum bars, a new heatsink will be used, in which ribs have appeared.

True, the cuts that form the fins are made only on one of the two aluminum bars. The second bar, connected to the first heat pipe, retained its simple shape. However, it is understandable. If ASUS engineers had endowed it with a developed structure, there would not have been a worthy place for the branded inscription “Rampage VI Apex”. And without it, the board would not have a chance to become a popular product.

For those who find the appearance of 12 cuts in one of the two heatsinks not enough, ASUS offers a special file frame with the board, which is put on the heatsink and allows you to mount an additional fan with an impeller diameter of 50 mm.

To confirm that the changes made to the shape of the heatsink really free the board from overheating, Der8auer overclocked the Core i9-7900X on it to 4.9 GHz. Without the use of a fan, the temperature of the VRM reached 103 degrees, but this did not cause throttling, which turns on when the power circuit reaches a temperature of 110 degrees. Thus, ROG Rampage VI Apex should be the first second-generation LGA2066 board that is not threatened by overheating of the power converter. And, apparently, it will be possible to overclock 12-, 14-, 16- and 18-core Core i9 processors on it, but this will already have to be done with the mandatory installation of an additional fan.

At the same time, the ASUS solution cannot be called elegant. The manufacturer did not change the design of the board itself and left the original eight-channel processor power scheme unchanged. And this means that all the improvements are just a kind of "crutch" that allows you to deal with overheating with the simplest methods.

Der8auer's full video about the new version of ASUS ROG RAMPAGE VI APEX:

The main component of any computer is the motherboard. It is to it that all the constituent parts of the system are connected, without it the functioning of any personal computer is impossible. Therefore, in the event of a malfunction, the user is faced with a lot of problems, the most basic of which is overheating. So, and how to determine the source of trouble?

The first signs of overheating and its consequences

The most common symptoms that the motherboard is heating up the following:

• A sharp reboot of the system, regardless of the user's occupation;
• The computer turns off by itself;
• Games freeze, crash, the system freezes when playing (only restarting the system helps).
• After 15-20 minutes after turning on the computer freezes.

List of components that are most often subject to overheating:

• Video card;
• CPU;
• Power Supply;
• Chipset on the motherboard;
• HDD.

In order to understand which of these overheats in one case or another, we will analyze each of them in more detail. But first, you should know with what programs you can determine the temperature of each of the components.

Motherboard Diagnostics

To get data on the temperature of all elements of the system, including the motherboard, you need to use convenient programs that you can easily find on the Internet. For instance, CPUID HWMonitor , or AIDA64. To monitor the temperature of the processor, you can use the program coretemp, for video card TechPowerGPU— Z. After installing these utilities, you can easily find out everything about the state of your system, voltage, temperature and more.

In most cases, the temperature norm for the motherboard can be considered 30-50 degrees, 60 is the limit. Often this is true, and these are normal indicators. For each individual plan, the instructions indicate its operating temperature range. We strongly recommend that you clarify this indicatively, since the ranges can vary significantly in different board models.

Modern motherboards are equipped with temperature sensors, but there are those on which they are absent. In this case, even with the help of programs, you will not determine the indicator of interest in any way. It is still possible to carry out a check, but only with the help of external thermometers, for example, an infrared non-contact thermometer.

The problem of overheating is most common in the warm season, when the air temperature begins to exceed 30 degrees, on hot days 35-40. Of course, this one has a very strong effect on your system, many gamers have problems with overheating even on absolutely working systems, only air conditioners and fans save them. Since, due to overheating, you can easily encounter a series of problems, and for online gamers, this is a disaster. If the motherboard does not survive severe overheating, it can burn the user most of the other components, almost always the processor “dies” immediately. A video card can also follow him.

Motherboard problems

For the average PC user, testing a motherboard is no easy task. But still, it is worth considering some signs that the computer is unstable precisely because of the motherboard.

The first thing you might think is that the board dies, and this, indeed, happens quite often. This process is irreversible, and it is better to immediately discard the idea of ​​repair, because even after it, its service life will remain unknown. It can work for a year or a week, but you will pay money for soldering or replacing components. If your PC is not yet quite an old man, it is best to buy a new board, preferably exactly the same if it suits you. If your computer is over 5 years old, you should consider a major upgrade. But this option is financially unprofitable, since at the current dollar rate, you can buy a car for the price of components on a PC.

Another reason is microcracks in the soldering of the board. In this case, it is better to immediately buy a new board and forget about the problem.

Does the computer turn off almost immediately after being turned on? The problem with this may be the failure of one or more capacitors on the motherboard. This case is very tolerable, since they are easy and simple to solder. It’s better not to bother with soldering yourself, since capacitors are different. It is better to take the PC to a service where they will not charge you a lot of money for this procedure.

Problems with the cooling system

The processor needs good cooling, and the more powerful it is, the more serious it should be. For example, an old-generation intelcorei3 needs a regular boxed cooler, but if you take one of the top-end i7 6900K, you won’t get off with a weak cooler, usually water cooling is installed on such processors. Also with regards to video cards, the cooling of a powerful video card must necessarily be active, that is, equipped with a radiator and a cooler, and preferably two or even three. Overheating of these components may be due to a cooling problem, perhaps one cooler stopped working, or it caught on a wire. Situations are different. But what should always be the same is the stable operation of the cooling system, and this applies not only to individual components, but also to the system unit as a whole.

Poor ventilation

The motherboard can easily get warm if the coolers are installed incorrectly in the case or they are missing. Air flows should be directed to the back of the case, thereby blowing out all the heat from the inside, plumes should not interfere with this, so they must be laid correctly, fortunately - modern cases make it possible to do this.

Ideally, there should be several coolers, one for blowing in, the other for blowing out. Thus, the air circulates evenly in the case, cooling all system components. If one of the coolers is not working well, there is no need to lubricate it, it is much easier to buy a new one.

When assembling the PC, you should not rush, all components must be properly fixed, the board must be fixed firmly and firmly on the case, all wires must be fixed so that they do not interfere with air circulation. By the way, from the visual part, it looks much more aesthetically pleasing, most likely, having seen your order in the system manager, many friends will be very surprised at this and think about the same for themselves.

Dust in the case

Once every few months, it is imperative to pay attention to dust accumulations in the system unit. The radiator collects dust very strongly, it also collects on the cooler blades, on the board itself. Many people think that dust is completely harmless, but it is not. Its large accumulations significantly impair the operation of the cooling system. Of course, because of this, the board and all parts of the system heat up more. An ordinary user will not be able to fully clean the computer from dust, we advise you to contact the service.

Radiator installed incorrectly

Does the board get hot? Does the computer turn off by itself? Stuck under load? Check if the heatsink is properly attached to the processor! The heatsink should fit very tightly to it, and a layer of thermal paste must be applied for better heat dissipation.

If the radiator is not mounted correctly, you can forget about the normal operation of the PC. The same will happen if the heatsink is not fixed correctly on the chipset of the board, on the video card. What kind of cooling can we talk about if the most important part of the cooling is not adjacent to the component cover?

Processor defective or defective

There were times when the board gets very hot, but the cooling system works fine. This may be due to a defective processor, or it has worn out over the years of use.

Also, one of the options for overheating can be to increase the voltage in the power supply, in which case it needs to be replaced and a voltage regulator should be bought. Unstable power operation can cause many malfunctions, so a high-quality and branded power supply should be taken.

Sometimes the problem of overheating can be in the wrong BIOS settings. With all this, specialists in the service center will easily figure it out. Without knowledge of the computer - self-repair is not worth it!

It all started with the fact that I started the long-planned alteration of the CBO circuit and temporarily transferred the percent and video to the air. Here is such a map
And shortly after that, I first encountered the notorious black screens, which you will agree with the 760pech is completely uncharacteristic.
Then I played exclusively in tanks and regular sorties from the battle infuriated incredibly.
Lowering the frequencies and increasing the voltage did not affect anything, but the latter helped to reveal an interesting feature: the card turned off not only in the maximum load mode, but also immediately after the load was dropped ...
It took me a while to figure out exactly what that meant.
Unfortunately, at that time there was nothing to measure the temperature with, since the old reobas with a temperature sensor had long been covered, and I still could not get together to buy a new one. Then the crisis hit and it was already completely not up to it ...
After about a week, I got tired of it, I removed the old Radeon 7950 from the farm for replacement - in order to calmly deal with the asus.
At the open stand, everything was the same as in the main computer. By that time, I already understood that somewhere and something was heating up, but how to determine where exactly? increase the airflow in order to test the theory of overheating, so to speak from different angles.
Luckily, I got my hands on the Scythe Kaze Server a few days later. Well, the first thing I began to figure out what was happening to me with the damn asus.
I put one sensor between the thermal pad and the heatsink of the power supply, and sent the second one to measure the temperature of the memory on the reverse side.

And the very first launch clearly illustrated that "something is wrong in the Danish kingdom ..." and specifically, some kind of indecency is happening!
Under load (WOT), the temperature of the BPM (radiator) quickly flew over 90g and continued to grow, demonstrating that this is clearly not the limit. the picture was taken before reaching the maximum temperature And by the time we left the game, there were even more...
And this is only on the radiator, and I don’t even want to imagine how much was on the mosfets at that time, but it’s a no brainer that more than 100, since thermal pads do not transfer heat well and the temperature drop easily reaches 10 or more degrees!

But then it’s even more interesting: after leaving the game, the temperature dropped slowly, slowly, and only after 3 minutes it dropped to 50g!
Hmmm, good gentlemen, you can’t live like that!.. I said to myself, mentally rolling up my sleeves and remembered another quotation that fits the spot: cut to hell! ©.

... finally, the disassembled card lies on the table, and I look at the BPM heatsink from a well-known manufacturer and quietly oh ... I roar:
no, I certainly assumed that everything was not in order there, but I did not think that everything was so bad.
So, the radiator ... best of all, my impressions could be conveyed by one or another obscene bend, but I will try to do without extreme sports, especially since some time has already passed and I have cooled down somewhat.

I must say right away that I have not yet seen a more crooked base - a strongly concave, wavy surface, simply killed with its absolute curvature, and the beginning of grinding clearly demonstrates this:
A squint-eyed Australopithecus cutting down this radiator with a stone ax stood before my eyes ... Damn it! ..
The measurements of the base showed that the difference between the edges and the center is almost 1.5mm! And this despite the fact that the thermal pad has a thickness of about a millimeter!
Therefore, before leveling with sandpaper, I had to attach a hammer.
Below are a few shots of the surface leveling process.

I had to level for a long time ... a very long time, but sooner or later everything ends. Finally, the base lost the last fragments of black paint, I polished the base a little more and washed it and sent it to dry.
While the radiator was drying, I began to think about how I should put it. The fact is that I categorically did not want to use a thermal pad, both old and new, but since the good ASUSopiteks soldered decently protruding capacitors between the mosfets, I needed to figure out how to get around this limitation.
Yes, in fact, it was not particularly necessary to invent - I already solved a similar problem and, by the way, also on the map from the asus. In short, I made a BPM matrix out of cardboard with parts that interfered with me and, grabbing a dried-up radiator, went to grind recesses for them. More precisely, I wanted to grind (with a dremel), but then I just drilled out the recesses.

Then, in order to guarantee not to arrange a short circuit, I applied my proprietary technology described in the note “hot-melt adhesive thermal pads”. The work is not at all tricky: having filled the recesses with hot-melt adhesive, I put a piece of thin film on the VRM and screwed the radiator. Here in the picture the radiator is installed on the film.
After the hot glue dried, I got a radiator with insulated recesses for protruding parts that simultaneously center the radiator during installation - beauty!

Well, for a snack, I was not too lazy to improve the cooling of the memory: all the same, the backplate blocks access to air and prevents at least some airflow. In general, it is completely incomprehensible what prevented it from being used as a heatsink for memory chips.
To implement this, I made cutouts in the backplate film opposite the memory I found aluminum plates in the bins and cut out 4 pieces according to the size of the memory and glued them to the backplate with hot glue so that they touched the memory chips. Here in the photo both the plates and the microcircuit are smeared with thermal paste before assembly.

And here is the result managed to bring down at least 20g on BPM and more than 10 on memory, which in my opinion is just fine!

After the work was done, no glitches were noticed! Which is exactly what was required!