• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

overheating issue

what cooler model it is? check also cooler fixing plate, screws tightening torque etc.! well, the base looks very cheap and not efficient...
This is a Deepcool AG400 WH, I checked all the screws and made sure they were perfectly set up, it seems pretty alright to me.

Have you checked that the planes of the cooler and processor are even throughout, without bends or depressions?
Yeah, nothing like that.
 
This is after cleaning, by the way. Tell me if I did a bad job

That cooler looks awful, it seems someone hand painted it white which may affect it's cooling capabilities....
Also , just 4 heatpipes..

Also your case is practically closed on the front side, if you want to see lower temperatures you'll need to replace the front panel with an open mesh one or just leave front panel off.
 
Last edited:
lmfao, are you using stock trash cooler? to get this thing heating to 80-s, you need massive overclock, which is again impossible due to poor chips binning lol...


what cooler model it is? check also cooler fixing plate, screws tightening torque etc.! well, the base looks very cheap and not efficient...
Yes, the 5500 comes with the Wraith Stealth Cooler which is even smaller than the Wraith Spire. That´s why I went for a AIO WC. Was going to upgrade to the 5800x3d but other things came first...
Anyways, the OPs cooler lookes like crap... Bet you don´t get much better temps with that cooler and the stock one.
Those naked heatpipes may not have good contact with the processor IHS. Keep in mind the bendings and irregularities.
This is what I mean: https://www.igorslab.de/en/cpu-heat...asures-now-at-amd-and-intel-rethink-basics/3/
Also, what´s your base room temp. Where I live it´s 10AM and it´s 25°C outside.
Cheers
 
That cooler looks awful, it seems someone hand painted it white which may affect it's cooling capabilities....
Also , just 4 heatpipes..

Also your case is practically closed on the front side, if you want to see lower temperatures you'll need to replace the front panel with an open mesh one or just leave front panel off.
I can't deal with the case right now, but if you want me to purchase a new cooler, I'd be glad to. What do you suppose is the best option for an R5 5500?
 
I can't deal with the case right now, but if you want me to purchase a new cooler, I'd be glad to. What do you suppose is the best option for an R5 5500?

Noctua 5-heatpipe NH-U12S or the NH-U12A with 7 heatpipes.
Thermalright and be- quiet has great coolers too but you may not be able to access your system memory with these coolers..
 
Noctua 5-heatpipe NH-U12S or the NH-U12A with 7 heatpipes.
Thermalright and be- quiet has great coolers too but you may not be able to access your system memory with these coolers..
i saw them. theyre great but man theyre ugly imo :(. should i maybe think of getting an AIO?
 
i saw them. theyre great but man theyre ugly imo :(. should i maybe think of getting an AIO?

A problem is, your case... There is no airflow from the front... To get lower temps you'll need a mesh front or remove the front panel..
 
your case has no ventilation. your cooler can work only if its able to pass normal air over the fins.

looks like you're just recirculating hot air inside the case...
 
I also agree that its most likely the lack of airflow, any half decent 4 heat pipe cooler should be able to handle that CPU.
I've had a Be Quiet! Pure Rock Slim on my 1600X for almost 4 years with no issues and thats a weaker cooler in overall in a In Win 101C case with 3 bottom intake and 3 exhaust fans.

As others suggested, you could try to take off the front panel and check if its any better just to be sure.
 
Sounds like your temps are normal for the bios settings you're using. Turn off bios PBO and download ryzen master and use PBO through it to potentially shave off a few degrees. Also try buying a better quality thermal paste, a lot of people go cheap with their paste because they think it doesn't matter but it does. Also try putting 1 fan in the top of your case in exhaust because your heatsync might be pooling hot air around your cpu. Tbh though the only way you'd really get noticeably lower temps would be to buy an AIO. In my opinion air coolers are dead and dated and only serve a purpose these days if you have a workstation system that has to be pulled apart all the time for maintenance, or if you took a 90's pc building course. I wouldn't recommend a custom loop as it's overkill for the CPU you have.

Try alphacool apex thermal paste if you can get it.

Arctic liquid freezer 2 is an excellent AIO for ryzen cpu's
 
Last edited:
In my opinion air coolers are dead and dated

Not really, a lot of quality air coolers beat AIO's, or are as good as.
It seems you never heard of Noctua, Be Quiet! or Thermalright air coolers.
 
Last edited:
Has anybody asked if bios has been updated to latest stable version?
 
Alright, I used ryzen master on 'auto overclocking' and let it perform the stress test, surprisingly unlike usual (where my cpu would instantly jump to 90c and slowly increase till the pc crashed), it only ever got up to around 78c. i dont know if that's a win or not, cause i haven't benchmarked my processor to see the performance yet. How can I do that?

also, scratch what i said about being a beginner, i'm definitely a complete noob :p
I'm curious about your PBO settings. Can you give us a screenshot of Ryzen Master?
 
I'm curious about your PBO settings. Can you give us a screenshot of Ryzen Master?
sure.
after some more tinkering and a repaste of my cpu, now my cpu temperatures are actually really low most of the time and my computer doesn't get very loud during gaming, so there's a chance the problem's either gone cause of one or the other.
1699772388788.png


Has anybody asked if bios has been updated to latest stable version?
I'm not trying to be that guy, but how would that help? I'm pretty sure my bios is about a year and a half old. Does it have to do with fan curves or anything?
 
I'm not trying to be that guy, but how would that help? I'm pretty sure my bios is about a year and a half old. Does it have to do with fan curves or anything?
A bug in the bios could for example cause the CPU to get too much voltage delivered, even on stock settings = more heat.
 
A bug in the bios could for example cause the CPU to get too much voltage delivered, even on stock settings = more heat.
Could we somehow check this without performing a bios update? I'm a very klutzy person and might mess it up, and I feel like simply checking the settings could tell us if that is the case.
 
Could we somehow check this without performing a bios update? I'm a very klutzy person and might mess it up, and I feel like simply checking the settings could tell us if that is the case.
You could go to the bios, check the bios version that is installed and as google about the bios version and known issues (and bugs) with it.
Youtube video about updating your mobos bios with usb stick:
Asus Prime B450 Plus Bios Update Tutorial Ryzen CPU 2019 - YouTube.

You can also try updating your bios over the internet inside the bios, so no need for any usb stick mess
 
cpu temps have nothing to do with a bios version, power limits or a so hot heatsink that opening a browser hits high 80s. it's basic physics.
and it's not like you run a 14900k with that cooler. this CPU can be cooled with sneezing on the IHS.

either it's a bad mount, loose mounting hardware on the board, not enough mounting pressure, bad paste, something that prevents proper heat transfer.
and having massive spikes with low loads indicates weak contact to the heatsink.
 
cpu temps have nothing to do with a bios version, power limits or a so hot heatsink that opening a browser hits high 80s. it's basic physics.
and it's not like you run a 14900k with that cooler. this CPU can be cooled with sneezing on the IHS.

either it's a bad mount, loose mounting hardware on the board, not enough mounting pressure, bad paste, something that prevents proper heat transfer.
and having massive spikes with low loads indicates weak contact to the heatsink.

And what was all the fuss about new AM5 Ryzens that were provided too much voltage not long ago? Burned out sockets and fried CPUs too. That was fixed with Bios update.
 
sure.
after some more tinkering and a repaste of my cpu, now my cpu temperatures are actually really low most of the time and my computer doesn't get very loud during gaming, so there's a chance the problem's either gone cause of one or the other.View attachment 321218


I'm not trying to be that guy, but how would that help? I'm pretty sure my bios is about a year and a half old. Does it have to do with fan curves or anything?
Whatever you did during the repaste probably fixed it. Like other said poor contact may have been the problem.

( I've never seen TDC(SOC) or EDC(SOC) before. That must be a new option. )

Did you do any benchmarking after changing the CPU power limits?
Did you optimize your values based on that?

You might find some more useful information in this post as well

And what was all the fuss about new AM5 Ryzens that were provided too much voltage not long ago? Burned out sockets and fried CPUs too. That was fixed with Bios update.
That was AM5 regarding too high SOC voltage.
 
Last edited:
Did you do any benchmarking after changing the CPU power limits?
Did you optimize your values based on that?
Yep, I used userbenchmark (although I've heard a lot of hate for it.) the R5 5500 averages 90% on it, but my tunes got it up to a 95-98% apparently, so that's nice.
 
Yep, I used userbenchmark (although I've heard a lot of hate for it.) the R5 5500 averages 90% on it, but my tunes got it up to a 95-98% apparently, so that's nice.
UB is interesting in that if you are also RAM overclocking it can show you how it impacts CPU utilization at the same time.
 
UB is interesting in that if you are also RAM overclocking it can show you how it impacts CPU utilization at the same time.
Well, I overclocked my ram to CL15. It was at 16-20-20-38, and I got it down to 15-19-19-35, so could that have made a difference?
 
Well, I overclocked my ram to CL15. It was at 16-20-20-38, and I got it down to 15-19-19-35, so could that have made a difference?
In UB you may see the CPU slider pull to the left (worse) or right (better) compared to your base line test.

Here is an old example of mine that I was able to find. The same kit of memory OC'ed at the same clock speed but with different timings.
See the change in CPU utilization. If your timings suck your CPU utilization will go down.

What I found by observing the benchmark was sometimes the absolute tightest settings aren't always the best settings.
You will have to try it and see what works for you. UB probably isn't the best for well rounded performance benchmarks which is why you want to do other kinds of tests to have a more realistic picture if you are tuning your system to your liking.

1699899158730.png


Generally what I have found is the closer you can keep your timings (in nanoseconds) to your kits specification, while raising frequency, the better OC you will have. If you increase your frequency but can't keep the timings (at least) in relative proportion you will have a worse result.


Here you can see a starting history 4yrs ago where I go through stock ram settings starting at 93.9%
to 109% on my RAM but killed my CPU utilization in the process of tightening timings.

( I've never seen TDC(SOC) or EDC(SOC) before. That must be a new option. )
I figured out where these SOC values come from. If you have an APU these will show up in Ryzen Master. The Ryzen 5 5500 is an APU chip with the graphics disabled.
 
Last edited:
Back
Top