[SOLVED] CPU throttling down at low temps, is it the PSU ?

xxxbabyxxx

Reputable
Dec 4, 2020
138
17
4,595
Hello,

i7-10700
RTX 3070
ML240L V2 on the CPU
32 GB Ram
Note: Temps/Airflow etc never a problem, nothing is overheating or too hot at all.

My CPU Is throttling down to 0.79 Ghz during games and stress tests (OCCT) even when temps are only in 50's 60's etc.
My CPU isn't thermal throttling during stress tests, and even during heavy loads on a hot day whilst playing 2021 games on maxed out settings the most it might reach is 80c°.

I've tried every single tweak/youtube video out there, turned Cstate on/off, hyper threading on/off, unplug the PSU and reset the power by holding power button etc.

- I've read that it might be a PSU issue.. could that be the case for me or is there anything else I should try first?
PSU: KCAS Plus GOLD 850W https://aerocool.io/product/kcas-plus-gold-850w/

Usually for my other machines I have some decent CoolerMaster platinum ones and such, but when I got this computer I had to get it with this PSU (to get the 3070 for $500 back then in 2020 lol)
I've had this PSU running almost 24/7 for a year and a half.. could be the PSU finally causing these problems? or is there anything I am overlooking or should check before we replace it?

Here are's a video showcasing the graph and the throttles (Note: The video was sped up a bit)

View: https://www.youtube.com/watch?v=WAi6jwWedNc

If It's the PSU, I'm thiniking of replacing it with this: G800 Gold from CoolerMaster https://www.coolermaster.com/catalog/power-supplies/g-series/g800-gold/

Thank you.
 
Last edited:
Solution
Hello,

i7-10700
RTX 3070
ML240L V2 on the CPU
32 GB Ram
Note: Temps/Airflow etc never a problem, nothing is overheating or too hot at all.

My CPU Is throttling down to 0.79 Ghz during games and stress tests (OCCT) even when temps are only in 50's 60's etc.
My CPU isn't thermal throttling during stress tests, and even during heavy loads on a hot day whilst playing 2021 games on maxed out settings the most it might reach is 80c°.

I've tried every single tweak/youtube video out there, turned Cstate on/off, hyper threading on/off, unplug the PSU and reset the power by holding power button etc.

- I've read that it might be a PSU issue.. could that be the case for me or is there anything else I should try first?
PSU: KCAS Plus GOLD...
Hello,

i7-10700
RTX 3070
ML240L V2 on the CPU
32 GB Ram
Note: Temps/Airflow etc never a problem, nothing is overheating or too hot at all.

My CPU Is throttling down to 0.79 Ghz during games and stress tests (OCCT) even when temps are only in 50's 60's etc.
My CPU isn't thermal throttling during stress tests, and even during heavy loads on a hot day whilst playing 2021 games on maxed out settings the most it might reach is 80c°.

I've tried every single tweak/youtube video out there, turned Cstate on/off, hyper threading on/off, unplug the PSU and reset the power by holding power button etc.

- I've read that it might be a PSU issue.. could that be the case for me or is there anything else I should try first?
PSU: KCAS Plus GOLD 850W https://aerocool.io/product/kcas-plus-gold-850w/

Usually for my other machines I have some decent CoolerMaster platinum ones and such, but when I got this computer I had to get it with this PSU (to get the 3070 for $500 back then in 2020 lol)
I've had this PSU running almost 24/7 for a year and a half.. could be the PSU finally causing these problems? or is there anything I am overlooking or should check before we replace it?

Here are's a video showcasing the graph and the throttles (Note: The video was sped up a bit)

View: https://www.youtube.com/watch?v=WAi6jwWedNc

If It's the PSU, I'm thiniking of replacing it with this: G800 Gold from CoolerMaster https://www.coolermaster.com/catalog/power-supplies/g-series/g800-gold/

Thank you.

That PSU looks good - the other thing that can trigger thermal throttling is if the motherboard VRM is overheating (I've run into this in the past on an old AMD FX based system).

Looking at your sig... you are running that 10700 on a H410 based board? If so that is likely the problem as the VRM solution on most H series boards is pretty weak. It's good enough for low end parts (Celeron, Pentium and i3 series processors) but it's asking quite a lot to run a 10700 on it...

The other issue is you are running with a liquid cooler - this is great for the CPU but unfortunately doesn't blow any air over the VRM components on the motherboard. As a test you could try opening the side of the case and pointing a fan over the motherboard around the CPU (the VRM components are usually positioned just around the edge of the CPU socket). If that improves your performance then that will be the issue. Your options are to either look at improving the airflow in your case and / or looking into a higher end motherboard with a stronger VRM.
 
  • Like
Reactions: xxxbabyxxx
Solution

xxxbabyxxx

Reputable
Dec 4, 2020
138
17
4,595
That PSU looks good - the other thing that can trigger thermal throttling is if the motherboard VRM is overheating (I've run into this in the past on an old AMD FX based system).

Looking at your sig... you are running that 10700 on a H410 based board? If so that is likely the problem as the VRM solution on most H series boards is pretty weak. It's good enough for low end parts (Celeron, Pentium and i3 series processors) but it's asking quite a lot to run a 10700 on it...

The other issue is you are running with a liquid cooler - this is great for the CPU but unfortunately doesn't blow any air over the VRM components on the motherboard. As a test you could try opening the side of the case and pointing a fan over the motherboard around the CPU (the VRM components are usually positioned just around the edge of the CPU socket). If that improves your performance then that will be the issue. Your options are to either look at improving the airflow in your case and / or looking into a higher end motherboard with a stronger VRM.
If it's the VRMs is there anything I can do other than get a new motherboard? I've had this system for a year and half only recently I've started to notice these problems :/

Note: I did open the glass side panel and put a strong fan right infront of it blowing strong, the results in OCCT were kind of the same, need to test it on other programs/games that had these problems.

What's intresting is usually in certain games, this drop in Ghz only happens at the beginning of the game, but never happens during the rest of the game, not sure if that means anything

I'm not sure as my VRM is below 100 (usually its high 80's or 90+ during tests) but it might've improved the minute i took the side glass panel off.. although that's just for games, when it comes to OCCT that isn't enough.

Is there any trick or anything I can do to make VRM heat less of a problem (outside of side panel off as im currently doing?)? so I can be 100% sure the problem is gone.
It's kind of weird though because I've had the same setup for more than a year now, and if anything this room itself got so many ventilation added to it compared to last year
 
Last edited:
I really doubt that the psu is the issue.

What is the make/model of your case, and what is the fan arrangement?
Can you show a pic of the radiator mount?

It is strange that throttling happens early and not later.
Look at your fan profiles.
One explanation could be that the fans start out too slow allowing heat to build before ramping up to increase cooling.
Is the pump speed set to 100%
 

xxxbabyxxx

Reputable
Dec 4, 2020
138
17
4,595
I really doubt that the psu is the issue.

What is the make/model of your case, and what is the fan arrangement?
Can you show a pic of the radiator mount?

It is strange that throttling happens early and not later.
Look at your fan profiles.
One explanation could be that the fans start out too slow allowing heat to build before ramping up to increase cooling.
Is the pump speed set to 100%
Thank you for replying,
do you mean it's strange that it showed this a year late? (not sure if you misswrote)

As for my case: https://aerocool.io/product/quartz-rgb/ (I know it isn't the best with front being glass instead of mesh, but overall GPU/CPU temps were always decent, like my GPU is usually 60's if I was to be mining or gaming or so, CPU temps also never been a problem. I used to see that VRM temps are okay to be around 100's or 105 or so.. so when mine were in the 80's and low 90's i thought it isn't the issue, but not sure it probably is at this point.. :/

As for my Radiartor, I have the fans blowing air OUT, mounted on top, so its exhausting hot air out of the case. Pretty much like this setup:
main-qimg-f93741d6a23b0db15264db35d5ec6df1

I've had this issue, updated my BIOS to F5 (which had Resizeable bar support), and I just re-updated it again to see if that helps, will check fan settings and let you know. Should I change anything?
I worried of keeping my AIO on 100% since I keep my pc on 24/7 and I often do Idle Mining on this machine when im asleep or so.

If I'm going to just use the PC without side panel, do you suggest reversing the FAN IN THE BACK to be intake then? I will have a house fan on the outside occasionally rotating and hitting the MOBO with some cool air every like 8 seconds or so (just a casual room fan that rotates)
 
Last edited:
Normally, throttling happens later when heat builds up.
Looking at your diagram, I note that the radiator tubes are at the top of the case.
Air, over time will eventually intrude into any aio system.
The bubbles will migrate up, collecting at the top.
As a theory:
When you start the system there will not be good fluid flow because of the bubbles at the top. In time, the pump should get fluid flowing properly.
But, until it does, cooling might not be as it should be.
It seems to me that a better orientation of the radiator would be to have the entry/exit tube location at the bottom.
AIO coolers do not last forever.
I think you could easily use a decent air cooler.
Here is a list of suitable noctua coolers for the i7-10700:
https://ncc.noctua.at/cpus/model/Intel-Core-i7-10700-413
If available height is an issue, look at the new NH-D12L
 
Last edited:
  • Like
Reactions: xxxbabyxxx

xxxbabyxxx

Reputable
Dec 4, 2020
138
17
4,595
Normally, throttling happens later when heat builds up.
Looking at your diagram, I note that the radiator tubes are at the top of the case.
Air, over time will eventually intrude into any aio system.
The bubbles will migrate up, collecting at the top.
When you start the system there will not be good fluid flow because of the bubbles at the top.
It seems to me that the proper orientation of the radiator would be to have the entry/exit tube location at the bottom.
AIO coolers do not last forever.
I think you could easily use a decent air cooler.
Here is a list of suitable noctua coolers for the i7-10700:
https://ncc.noctua.at/cpus/model/Intel-Core-i7-10700-413
If available height is an issue, look at the new NH-D12L
I've ran few tests right now on the games that I play with side panel off, and room fan rotating, it appears as if the issue is gone, but if I was to hit it with something tough like OCCT then i'd notice it, but for the programs/games that I use everyday it appears to be fixed/gone, would that be okay for my case to ignore it now and just avoid using side panel during the summer?

Oh yeah also not sure why it does that, I tried turing C State off, my CPU stays around 4.59 Hz on idle and such, If I play it might drop a bit but atleast it's not spiking down to 0,79 Ghz like it was before..

On OCCT its still spiking that low, but its taking few extra seconds before it does the spikes, where before the spikes were almost instant
 
Last edited:
OCCT is a stiff test.
Probably not at all what your normal work is doing.
Try running the CPU-Z stress test.
If your windows power profile is performance, it will run the cpu at max all the time.
I think you want to select balanced.
Set a minimum cpu% to something like 20%
That way, the cpu will lower the multiplier when there is little to do.

It does not look like you have the option to mount the radiator on top, sending hot air out of the case. If you could, that is what I would do.
By leaving the side cover off, your parts will accumulate dust more quickly.
I would try to avoid that if possible.
Keep the rear fan as exhaust.
You want to get heat out of the system as fast as possible.
 

xxxbabyxxx

Reputable
Dec 4, 2020
138
17
4,595
OCCT is a stiff test.
Probably not at all what your normal work is doing.
Try running the CPU-Z stress test.
If your windows power profile is performance, it will run the cpu at max all the time.
I think you want to select balanced.
Set a minimum cpu% to something like 20%
That way, the cpu will lower the multiplier when there is little to do.

It does not look like you have the option to mount the radiator on top, sending hot air out of the case. If you could, that is what I would do.
By leaving the side cover off, your parts will accumulate dust more quickly.
I would try to avoid that if possible.
Keep the rear fan as exhaust.
You want to get heat out of the system as fast as possible.
I've ran a CPU Z Stress test as you said, (I had discord, firefox, battle.net and few other programs alongside manually recording)
Here's the results (less than a minute test):
unknown.png



unknown.png


1 - CPU Temps were dropping down to 40c, it dropped but it wasn't as bad as OCCT ones, those dropped it down to 0.79 Ghz
2 - I noticed throttling once VRMs reaches 96-98c

As for mounting radiator on top, I do have it there currently, sending hot air out of the case, and the rear fan is exhausting ( I adjusted it in BIOS to run full speed, previously it was running too slow for some reason 800 rpm)
 
@xxxbabyxxx
Run ThrottleStop and open up the Limit Reasons window to find out why your CPU is throttling.

This sounds like a BD PROCHOT throttling issue. A sensor, external to the CPU, might be sending a false or premature throttling signal to the CPU. This is what usually causes sustained drops to 0.79 GHz.

Normal thermal throttling in Intel 10700 CPUs does not start until one of the cores reaches 100°C. During normal thermal throttling, the CPU will continue to run as fast as possible. It will stay pegged at 99°C to 100°C the whole time. If none of the cores are anywhere near this temperature and you are getting severe throttling then BD PROCHOT is almost always the reason why.

ThrottleStop has a BD PROCHOT box on the main screen. Clearing this box tells the CPU to ignore any external throttling signals. Your CPU will still thermal throttle if it ever gets too hot. ThrottleStop does not have any features to turn off that safety feature.

If this is the problem, the only fix is a new motherboard or you will need to always run ThrottleStop. Running a program that costs nothing seems like the best solution to me. You can use the Task Scheduler to add ThrottleStop to your Windows start up sequence and you will not have to worry anymore about some broken sensor that has gone bad.

If Limit Reasons shows VR throttling, that one cannot be ignored. You need to put a fan pointing towards the voltage regulators to try and cool them down.
 
  • Like
Reactions: xxxbabyxxx

xxxbabyxxx

Reputable
Dec 4, 2020
138
17
4,595
Testing with Throttlestop by disabling BD PROCHOT actually caused my system to shut down during a stress test, I guess because the VRM temps were maybe really high and the system shut itself down? so I guess I won't use that program anymore just to be safe..

Here's what's intresting though, upon googling BD PROCHOT I saw that majority of threads were posted about people complaining about LAPTOPS.. I want to ask ,could a registery cause this?

I'm on a Desktop but few months ago I did add 2 registeries to make my PC able to access certain programs in store, the registery set were "Dell Inspiron" and when I googled now it seems like its a laptop.. I had deleted those registery but i am not sure if somehow its causing something on my vrm/cpu to behave different or if it's relevant at all to BD PROCHOT?
 
I guess because the VRM temps were maybe really high
ThrottleStop Limit Reasons should report in red VR TEMP if the voltage regulators are too hot. It will report VR CURRENT if the voltage regulators are exceeding their current capability. The CPU will then throttle based on this. It should not just shut down because of the voltage regulators.

What stress test were you running? Lots of computers are not 100% stable at default settings during some stress tests. The H410M is a low end motherboard. It might not have what it takes to run a 10700 at full speed during a stress test. These boards were probably more intended for 10400 buyers.

There are extreme stress tests like Prime95 Small FFTs that a lot of computers are not capable of running reliably. When your computer crashes, you cannot just assume that it must be the voltage regulators causing the problem.
 
  • Like
Reactions: xxxbabyxxx

xxxbabyxxx

Reputable
Dec 4, 2020
138
17
4,595
ThrottleStop Limit Reasons should report in red VR TEMP if the voltage regulators are too hot. It will report VR CURRENT if the voltage regulators are exceeding their current capability. The CPU will then throttle based on this. It should not just shut down because of the voltage regulators.

What stress test were you running? Lots of computers are not 100% stable at default settings during some stress tests. The H410M is a low end motherboard. It might not have what it takes to run a 10700 at full speed during a stress test. These boards were probably more intended for 10400 buyers.

There are extreme stress tests like Prime95 Small FFTs that a lot of computers are not capable of running reliably. When your computer crashes, you cannot just assume that it must be the voltage regulators causing the problem.
Thanks for replying,

I ran CPU Z Stress test for under a minute, CPU stayed on 100% was working fine, VRMs were 98c max, it went okay!
I ran OCCT Stress test for under a minute, it crashed within few seconds. (Basically shutdown completely, no blue screen or anything)

If the issue is related to BD PROCHOT, could be related to the Registery tweak I once did awhile ago changing my Desktop PC to "Dell Inspiron" for few days? which apperantly is a Laptop, could that have messed with any of the mobo or vrm? (idk if that spossible), it was just weird that I googled BD PROCHOT and most of them were laptop related then i remember I had that registery thing to make it seem as if my OEM is a dell laptop so i can access more microsoft store products

All I did was add 2 registery lines like this video said at 2:11
View: https://www.youtube.com/watch?v=eQHkIvJKegA


but what's weird I think it creates other registery for you as a DELL INSPIRON/OEM 2 tabs below, idk if it did anywhere else too :/
 
Last edited:
Your registry tweak and BD PROCHOT throttling have nothing to do with each other.

Use the Disable Turbo option in ThrottleStop to run your CPU at its bsse frequency. Can you run OCCT then without it crashing?

Can you run Cinebench at full speed without crashing? That is a more realistic stress test compared to OCCT.

If you make a marathon runner wear a 100 pound back pack on a 35C day and he keels over half way through the race, does that mean he is a bad runner? Not really. Buy a more expensive board next time if running OCCT is important to you.

Add a fan inside your case so there is air flow towards the voltage regulators.

You can use ThrottleStop to reduce the core and cache voltage by -50 mV. This can help reduce power consumption and heat.
 

xxxbabyxxx

Reputable
Dec 4, 2020
138
17
4,595
Your registry tweak and BD PROCHOT throttling have nothing to do with each other.

Use the Disable Turbo option in ThrottleStop to run your CPU at its bsse frequency. Can you run OCCT then without it crashing?

Can you run Cinebench at full speed without crashing? That is a more realistic stress test compared to OCCT.

If you make a marathon runner wear a 100 pound back pack on a 35C day and he keels over half way through the race, does that mean he is a bad runner? Not really. Buy a more expensive board next time if running OCCT is important to you.

Add a fan inside your case so there is air flow towards the voltage regulators.

You can use ThrottleStop to reduce the core and cache voltage by -50 mV. This can help reduce power consumption and heat.
Thank you for replying,

I'll do the Cinebench test today and let you know how it goes. Idk if CPU Z would've crashed because I waited around 10 seconds+ and it didnt spike, usually it immediatly spike, so I thought it worked (hence why i quickly went on to do OCCT in happiness where my pc shut down lol), but NOTE: my PC didn't shutdown with BD PROCHOT on during Warzone loading, which is where i always see that problem show up daily.. (Pretty much i see it daily there but that's the only program/software that bothers me, other than that nothing seems to spike or use that much CPU usage.. ) I do render videos once a month or every few months but usually rendering takes few seconds and isnt a problem.. I actually rendered few past few days but didnt track if its spiking

Would me trying Cinebench with BD PROCHOT turned off, would that damage my computer if i got another case of "system shut down"? is it damaging my pc that my pc is shutting down itself during these?
 
would that damage my computer
Not likely. It sounds like your computer shut itself off because your settings are not stable. The default BIOS settings are usually good enough so a computer can start up and run but these same settings may not allow your computer to run an extreme stress test like OCCT reliably.

Manufacturers that do not have time or money to figure out these problems will add things like BD PROCHOT throttling to cover up the mess that they created. A computer should not ever have to use BD PROCHOT throttling during normal use. It should never randomly shut down whether BD PROCHOT is enabled or not.

Lowering the CPU core and cache voltage -50 mV might help. This reduces power consumption and heat output. That will give your CPU some more headroom. Have you tried doing that yet?
 
  • Like
Reactions: xxxbabyxxx

xxxbabyxxx

Reputable
Dec 4, 2020
138
17
4,595
Can you run Cinebench at full speed without crashing? That is a more realistic stress test compared to OCCT.
Thanks for your replies,

I ran Cinebench and it behaved the same (although took sometime to start spiking), I closed it before it was done but it was almost over, CPU Usage was 100% and spiked down all throughout middle/late of the test.

unknown.png


It didn't crash though but that was with BD Prochet turned on, I haven't tried giving it less voltage yet, you recommend a reliable method/program for that? I tried to have C States on/off to see if it makes a difference but i think it was the same, so we know for sure that its VRM related right? since opening side glass panel and having a fan hitting mobo managed to remove the lag I had in COD: Warzone which i tried to fix for 3 weeks.

On that game, these spikes used to happen every single game loading, but currently it never does it anymore at all unless I'm stress testing it (occt/cpuz/cinebench).

And yeah I wish i had a better MOBO but it's what I could get back then, the VRM doesn't have a heatsink only the chipset does on that mobo.

Oh also, idk if you noticed but IS THIS NORMAL?
unknown.png


I believe it's been like this SINCE i got the computer years ago, but i am not entirely sure
 

Phaaze88

Titan
Ambassador
Oh also, idk if you noticed but IS THIS NORMAL?
unknown.png
I HATE HWMonitor for this.
Those are all motherboard sensors, but it's not specified which nor where, making them useless for the most part.
Some users confuse them for cpu core temperatures, because 'Temperatures' with a thermostat icon next to it.
Plus for some dumb reason, this set of sensors is laid out BEFORE the actual cpu core temperatures, which - get this - is also labeled 'Temperatures' with a thermostat icon next to it.

Yeah, that's definitely not confusing for those who don't know... /s
 
  • Like
Reactions: xxxbabyxxx

xxxbabyxxx

Reputable
Dec 4, 2020
138
17
4,595
I HATE HWMonitor for this.
Those are all motherboard sensors, but it's not specified which nor where, making them useless for the most part.
Some users confuse them for cpu core temperatures, because 'Temperatures' with a thermostat icon next to it.
Plus for some dumb reason, this set of sensors is laid out BEFORE the actual cpu core temperatures, which - get this - is also labeled 'Temperatures' with a thermostat icon next to it.

Yeah, that's definitely not confusing for those who don't know... /s
Right exactly it got me panicking :sweatsmile:

Well then guys, having glass side panel open, with my room fan rotating at me & the Pc completely solved my issue for all the daily stuff I used to do, like my CPU never stutter because of VRMs anymore, EXCEPT FOR when i do stress tests ( CPU Z, OCCT, Cinebench). I render videos maybe once every few months and it usually goes fast without a problem.

-
Do I still need to undervolt my CPU or can I just ignore the problem? I'm asking because my CPU is reaching 4.59 Ghz in games without spiking down to 0.79Ghz like it did before, because I got my glass panel open. I feel like if I undervolt it, it might do better in stress tests but not the daily tasks I do such as gaming and whatnot, what do you think?
 
Thanks for your replies,

I ran Cinebench and it behaved the same (although took sometime to start spiking), I closed it before it was done but it was almost over, CPU Usage was 100% and spiked down all throughout middle/late of the test.

unknown.png


It didn't crash though but that was with BD Prochet turned on, I haven't tried giving it less voltage yet, you recommend a reliable method/program for that? I tried to have C States on/off to see if it makes a difference but i think it was the same, so we know for sure that its VRM related right? since opening side glass panel and having a fan hitting mobo managed to remove the lag I had in COD: Warzone which i tried to fix for 3 weeks.

On that game, these spikes used to happen every single game loading, but currently it never does it anymore at all unless I'm stress testing it (occt/cpuz/cinebench).

And yeah I wish i had a better MOBO but it's what I could get back then, the VRM doesn't have a heatsink only the chipset does on that mobo.

Oh also, idk if you noticed but IS THIS NORMAL?
unknown.png


I believe it's been like this SINCE i got the computer years ago, but i am not entirely sure
Most of those are unused sensor inputs of the superIO monitoring chip.

The monitoring chip has capability of monitoring several more temp sensors than the mfr. has on the board...or chose to wire to the chip. Those readouts then sort of 'free-float', giving you random and sometimes extreme values that make no sense.
 
  • Like
Reactions: xxxbabyxxx
Right exactly it got me panicking :sweatsmile:

Well then guys, having glass side panel open, with my room fan rotating at me & the Pc completely solved my issue for all the daily stuff I used to do, like my CPU never stutter because of VRMs anymore, EXCEPT FOR when i do stress tests ( CPU Z, OCCT, Cinebench). I render videos maybe once every few months and it usually goes fast without a problem.

- Do I still need to undervolt my CPU or can I just ignore the problem? I'm asking because my CPU is reaching 4.59 Ghz in games without spiking down to 0.79Ghz like it did before, because I got my glass panel open. I feel like if I undervolt it, it might do better in stress tests but not the daily tasks I do such as gaming and whatnot, what do you think?

I would say that you are pushing your motherboard pretty hard given it is thermal throttling due to the VRM's. Undervolting won't harm anything although if you go too far it will make the machine unstable. It's worth trying as anything that takes pressure off the motherboard is going to improve the overall longevity of the machine. The main risk with leaving it as is, is that your motherboard will eventually fail due to running components close to the limit for long periods.

The other thing you could do is look at making a modification to improve cooling of the VRM components - you can add heat sinks to the VRM's using thermally conductive adhesive pads, and / or add a small fan to blow air over the VRM will help keep the temps much lower and would be safer for long term stability of the system.
 
  • Like
Reactions: xxxbabyxxx

xxxbabyxxx

Reputable
Dec 4, 2020
138
17
4,595
I would say that you are pushing your motherboard pretty hard given it is thermal throttling due to the VRM's. Undervolting won't harm anything although if you go too far it will make the machine unstable. It's worth trying as anything that takes pressure off the motherboard is going to improve the overall longevity of the machine. The main risk with leaving it as is, is that your motherboard will eventually fail due to running components close to the limit for long periods.

The other thing you could do is look at making a modification to improve cooling of the VRM components - you can add heat sinks to the VRM's using thermally conductive adhesive pads, and / or add a small fan to blow air over the VRM will help keep the temps much lower and would be safer for long term stability of the system.
Thanks for replying, right now I have a fan in my room that's blowing towards my desk and rotates towards the VRMs, on idle I think it's at 40-50c, on workload that uses the most of my cpu (COD Warzone) its reaching 87c maximum, during a hot day. So outside of stress tests those are the temps. the VRM only reaches 85+ during the part where the game is loading and after that it doesn't appear to be getting higher than 87c during few hours playing and recording etc.

So I guess just opening the side panel and have my fan rotate did magic for me.. I was worried at first thinking its the CPU not getting proper power from the PSU but I'm glad to know its something as simple as VRM temps

Thank you all for the replies <3 Tomshardware saves the day as always XD
 
Last edited:
  • Like
Reactions: uWebb429 and cdrkf