Iridium28 wrote:
I'd recommend starting with remounting your cooler. Take the CPU cooler off, clean up the thermal paste on both the CPU IHS and cooler base, reapply thermal paste and re-mount the cooler. Based on your description I'd lean toward this being the problem.

You can also try entering bios and disabling MCE (ASUS Multicore Enhancement). I'd doubt this would create a 22C increase in temperature for that processor at default bios settings, and it certainly shouldn't lead to 60C at idle, but it's easy to check temps with it disabled. You should be able to find that within the AI Tweaker tab in bios under advanced mode.

Thanks for your reply

The cooler is mounted correctlly and i desabled the MCE but temperatures seems the same. I don't know if the voltages by default are set correctly cuz i see temperature spikes on Core #1 and is hotter than the rest. Now i'm getting around 35º idle all cores except the core #1 39º/40º
By the way i'm buying a new non stock cooler this week to improve temps

Well, the cooler recommendation was because it would be the expected behavior if there was an issue there. Not enough thermal paste, too much, a poor spread, not enough mounting pressure, the cooler isn't seated perfectly, etc. This type of issue inevitably finds a way to happen when moving processors to new boards. It's also partially why it's generally a best practice to remove and reapply thermal paste when doing so.

If it's not an issue there I would start by updating your bios to the latest version.

If this doesn't fix the problem check voltages. Boards tend to default to higher voltages than necessary to ensure stability under auto settings (as do CPU's in terms of what they request from the board, or the VID). Some are worse than others about it. So if you can, try to find out where those are running. I'd recommend installing HWinfo for monitoring software to check this because it's quite robust and often more accurate compared to various alternatives. Specifically, find out what it is showing for the following at idle and when under load...

https://www.hwinfo.com/download/

1. Vcore
2. CPU_VCCSA
3. CPU_VCCIO

You might have to scroll down to find the values in the program. They should be listed under the heading for your board. While you're there, check fan speeds as well. They are likely under the same heading for the board. Verify those are running at the expected values when idle and under load.

I'd check what the bios is defaulting to for LLC (Load Line Calibration). You can probably find this under AI Tweaker > Digi+ VRM, or something similar. It should tell you what it's aiming for beside the setting. If it's displaying 7 or one of the higher values I'd manually drop it down to a middle of the pack value (around 3-4 on ASUS boards).

All of this aside, upgrading the cooler is probably wise. I do not believe I've ever come across a processor where the default cooler was.... satisfactory. 8700's like to behave as if they're trying to find out what it feels like to survive re-entry in the first place.

Iridium28 wrote:
Well, the cooler recommendation was because it would be the expected behavior if there was an issue there. Not enough thermal paste, too much, a poor spread, not enough mounting pressure, the cooler isn't seated perfectly, etc. This type of issue inevitably finds a way to happen when moving processors to new boards. It's also partially why it's generally a best practice to remove and reapply thermal paste when doing so.

If it's not an issue there I would start by updating your bios to the latest version.

If this doesn't fix the problem check voltages. Boards tend to default to higher voltages than necessary to ensure stability under auto settings (as do CPU's in terms of what they request from the board, or the VID). Some are worse than others about it. So if you can, try to find out where those are running. I'd recommend installing HWinfo for monitoring software to check this because it's quite robust and often more accurate compared to various alternatives. Specifically, find out what it is showing for the following at idle and when under load...

https://www.hwinfo.com/download/

1. Vcore
2. CPU_VCCSA
3. CPU_VCCIO

You might have to scroll down to find the values in the program. They should be listed under the heading for your board. While you're there, check fan speeds as well. They are likely under the same heading for the board. Verify those are running at the expected values when idle and under load.

I'd check what the bios is defaulting to for LLC (Load Line Calibration). You can probably find this under AI Tweaker > Digi+ VRM, or something similar. It should tell you what it's aiming for beside the setting. If it's displaying 7 or one of the higher values I'd manually drop it down to a middle of the pack value (around 3-4 on ASUS boards).

All of this aside, upgrading the cooler is probably wise. I do not believe I've ever come across a processor where the default cooler was.... satisfactory. 8700's like to behave as if they're trying to find out what it feels like to survive re-entry in the first place.

Thank you for your time and help bud
I already update the BIOS from the out of the box (2017) to the last updated one (2019) but no changes on temps
I will try all that you say when i get home and i will screenshot all stats of the HWinfo software.
The main problem is the board, i know that bocause i have another one but H370. Replasing the Asus Strix Z370 with my old Aorus H370 Gaming 3 drop temps considerably

With Gigabyte AORUS H370 Gaming 3:

Cold Boot BIOS Temperature: 32 degrees celcius
Idle Temperature: 33-36 degrees celcius
Gaming Temperature: 62-76 degrees celcius
Full Load / Stress Test Temperature: 80-100 degrees celcius


With ASUS Strix Z370 Gaming-F:

Cold Boot BIOS Temperature: 35 degrees celcius
Idle Temperature: 38-45 degrees celcius
Gaming Temperature: 70-90 degrees celcius
Full Load / Stress Test Temperature: 90-100 degrees celcius

Iridium28 wrote:
Well, the cooler recommendation was because it would be the expected behavior if there was an issue there. Not enough thermal paste, too much, a poor spread, not enough mounting pressure, the cooler isn't seated perfectly, etc. This type of issue inevitably finds a way to happen when moving processors to new boards. It's also partially why it's generally a best practice to remove and reapply thermal paste when doing so.

If it's not an issue there I would start by updating your bios to the latest version.

If this doesn't fix the problem check voltages. Boards tend to default to higher voltages than necessary to ensure stability under auto settings (as do CPU's in terms of what they request from the board, or the VID). Some are worse than others about it. So if you can, try to find out where those are running. I'd recommend installing HWinfo for monitoring software to check this because it's quite robust and often more accurate compared to various alternatives. Specifically, find out what it is showing for the following at idle and when under load...

https://www.hwinfo.com/download/

1. Vcore
2. CPU_VCCSA
3. CPU_VCCIO

You might have to scroll down to find the values in the program. They should be listed under the heading for your board. While you're there, check fan speeds as well. They are likely under the same heading for the board. Verify those are running at the expected values when idle and under load.

I'd check what the bios is defaulting to for LLC (Load Line Calibration). You can probably find this under AI Tweaker > Digi+ VRM, or something similar. It should tell you what it's aiming for beside the setting. If it's displaying 7 or one of the higher values I'd manually drop it down to a middle of the pack value (around 3-4 on ASUS boards).

All of this aside, upgrading the cooler is probably wise. I do not believe I've ever come across a processor where the default cooler was.... satisfactory. 8700's like to behave as if they're trying to find out what it feels like to survive re-entry in the first place.

Hi there, I installed HWInfo and this is what i got on Vcore, CPU VCCSA and VCCIO


I swap again to the Aorus H370 Gaming 3 and make new test, i got very good temperatures with that board.

Colder core - Hotter core / max Tem
BIOS 29º C
Desk Idle 29º/33º C
Gaming PUBG LITE 40º-47ª/61º C
Gaming PUBG STEAM 39º-45º/67º C
Gaming The Witcher III 51º-61º/ 60º-74º C
FULL LOAD with STRESS CPU-Z: Max core speed: 4300mhz / max temp: 85º (20 seconds long)
Min / Max overall: 28º / 85º

RoshGG wrote:
Hi there, I installed HWInfo and this is what i got on Vcore, CPU VCCSA and VCCIO


I swap again to the Aorus H370 Gaming 3 and make new test, i got very good temperatures with that board.

If changing the board causes your problem I'd lean toward the ASUS board being the culprit as well.

In terms of your HWInfo....

Vcore at ~1.25 as a maximum at auto doesn't seem that far out there. Your CPU could probably run it lower but I'd leave it alone for now.

VCCIO and SA could probably be a little lower but I would leave them alone for now (ASUS bios like to inflate these on auto...).

I will say your memory voltage looks off. You said you weren't using XMP and it's showing 1.52v on the ram. Assuming it's reading correctly that looks... really high. I think that memory is supposed to default to 1.35v while under XMP. I'd check your bios to make sure nothing is set incorrectly for the memory. One option there would be to load into bios and reset it to default settings. If that doesn't fix it you may need to manually assign it's voltage to 1.35 (can prob go lower toward 1.2v or so with XMP off entirely but to avoid any issues I'd start at 1.35v).

Based on the power draw in that HWInfo I'd keep MCE disabled. You've indicated you want to run the system at stock settings and MCE really... wouldn't be.

The only other thing I can think of is the CPU cooler mounting mechanism with that board isn't creating enough pressure between the cooler and CPU IHS.

Beyond that upgrading the default CPU cooler would probably go a long way toward improving your temps.

Iridium28 wrote:
If changing the board causes your problem I'd lean toward the ASUS board being the culprit as well.

In terms of your HWInfo....

Vcore at ~1.25 as a maximum at auto doesn't seem that far out there. Your CPU could probably run it lower but I'd leave it alone for now.

VCCIO and SA could probably be a little lower but I would leave them alone for now (ASUS bios like to inflate these on auto...).

I will say your memory voltage looks off. You said you weren't using XMP and it's showing 1.52v on the ram. Assuming it's reading correctly that looks... really high. I think that memory is supposed to default to 1.35v while under XMP. I'd check your bios to make sure nothing is set incorrectly for the memory. One option there would be to load into bios and reset it to default settings. If that doesn't fix it you may need to manually assign it's voltage to 1.35 (can prob go lower toward 1.2v or so with XMP off entirely but to avoid any issues I'd start at 1.35v).

Based on the power draw in that HWInfo I'd keep MCE disabled. You've indicated you want to run the system at stock settings and MCE really... wouldn't be.

The only other thing I can think of is the CPU cooler mounting mechanism with that board isn't creating enough pressure between the cooler and CPU IHS.

Beyond that upgrading the default CPU cooler would probably go a long way toward improving your temps.

Man, Thanks for all your time and help with this. I will test lower the ram voltage to 1.35v.
A friend of mine tells me to clear CMOS and see if thats do something.

On the other hand, i upload a video to youtube showing a CPU-Z Stress Test and HWInfo if you want to see it


I gonna buy a new cooler this week and test again 🙂

sry. double post.

Yeah if temps are boosting that quickly I'd lean back toward a bad cooler contact. HWINFo was only showing it up to 4.4-4.3ghz in your test and it's already 90C+. The test may not be stressing it enough to get your cores maxed. Regardless, 90C+ under stock settings below the max CPU turbo boost frequency indicates something is wrong. Voltages, or more specifically vcore, would have to be off by a decent margin to cause such behavior.

If it helps, I recently replaced a CPU cooler because it was creating the same problem. It would run a little high at idle and placing any stress on the CPU would immediately jack the temps up. In pretty much any stress test I'd hit thermal throttling damn near instantly. It was a cheap AIO I picked up at like 65% off at one point. I suspect the pump died in it. Replacing the cooler immediately solved the problem. It's not completely comparable to your situation but it illustrates what happens with a bad cooler or bad contact for it. Temps will skyrocket extremely quickly.

I'd revisit the cooler mount again. In some cases I've seen mention of the mounting mechanisms playing poorly with certain boards. In a way where everything appears seated properly but, for whatever reason, the cooler doesn't contact the chip correctly. You usually have to pair washers (of the non-conductive variety, ideally) with the mounting hardware to fix it. I'd think damn near every board would be built to play well with stock coolers but assumptions aren't safe.

I guess resetting bios is worth a try. The easiest way to do that would be to unplug the power cable from the wall or back of the PSU (flip off the PSU switch too if you want), connect/jump the two bios reset pins together for ~10 seconds and plug it in again. From your board manual it looks like the pins are right above your system panel header. I'd recommend using a screwdriver to bridge the two pins (no sense fumbling with one of those tiny jumpers.... not unless you enjoy scavenger hunts). Alternatively, if you're blessed with a bios reset button push/hold that for ~10s after unplugging the system (don't see one in the manual for your board).

The only other thing I can think of is the board is faulty. Software monitoring is nice but it's at best a ballpark. Furthermore, it's reliant on the applicable sensors. Those could be reading out incorrect values. It would conceivably be possible for your monitoring software to claim the board or chip is running at X voltage, or Y temperature but it's either doing something else or reporting it wrong. In other words, take the monitoring readouts with a grain of salt.

You said you were picking up a new cooler. If it is arriving within the return window for the board I'd install it and see if your problem fixes itself. If you're 100% certain your cooler is mounted correctly and/or the new cooler doesn't fix the problem I'd RMA the board, to be honest.