#My GPU is so hot, its at 80 Celcius, with 100 perc Utilization
1 messages · Page 1 of 1 (latest)
*notes:
-i have replaced the thermal paste on my GPU
-If i place everything at Ultra, my GPU becomes so hot, it loses signal to my monitor. This relates to the title of the thread.
Can someone suggest what GPU should i have so that i can reduce the GPU util on Ultra settings?
You could try using MSI Afterburner (free) and a custom fan curve. To be honest, I think it is considered normal for a GPU to use 100% utilization while playing a game; it will throttle fps, etc. if there is a problem. I use a 1080Ti with everything on Ultra and have not experienced problems with the GPU despite 100% utilization. I also don't believe that 80 deg. while playing is abnormal, particularly for games using the Apex engine (anything from Avalanche Studio Group; all their games tend to run hot). https://www.msi.com/Landing/afterburner/graphics-cards
Hello @dusty trench we appreciate your concern and would like to kindly ask you to submit a support ticket.
Could you please send us a message with the information contained in this thread and your dxdiag and game version ? That way we can investigate them, and if possible, resolve them. https://support.thehunter.com/hc/en-us/requests/new
GPU's running at 100% is normal. If it's below 100%, it means you have a bottlenecking problem.
yep. it is quite literally optimal to have your gpu reach about 100% utilization in games
that unless your framerates are being lowered by vsync or any sort of frame limiter
For now, i am currently reinstalling the game. I will try using a frame limiter for now. If it still cant fix my problem, i will lodge the report.
It sounds like the cooling system just isn't doing a good enough work. Can you improve it somehow?
Ill try with the frame limiter for now.
Frame limiter didnt work, now gonna see if changing the HDMI cables will do anything.
You should go get a fire truck
Cables didnt work.
I had lodge the report, although i dont know what DXGI is.
Get water cooling or just buy a new one
Try to limit your monitor hz
Sponsor me one, seems like ya have some money to say that to me breh
60Hz, i believe, dunno if i can reduce it more.
i wouldnt go below 60
Ill give it a try, if it doesnt work, then meh.
btw what 3060Ti do u have
i have one from gigabyte and its been a pain in the ass for me too
3060Ti Vision
i assume gigabyte also
I think most of the methods ive checked is not hardware not software, but i believe its down to the game itself, because i played on my sister's PC with the same graphics card and its working fine
I think its the graphics card. Gigabyte has had a lot of problems with its 3060Ti from what ive seen and heard on the internet, maybe try and contact the place youve bought it from and ask if they can send a replacement/troubleshoot it for you
I use 144
Not worth it
It is
I've set up my screen hz to 60 even if it can do 144, and I cant really tell the difference. My computer emits less heat tho, and that helps in spain
theres a huge difference between 60 and 144 i think youve done something wrong with your settings. youve either had it on 60 the whole time or somethinh
Nah, but the difference aint that much to bother the cpu with it
No advantages on imputting those frames, better let the lil dude be more relaxed so it doesnt burn
its not the cpu tho
Still the same.
honest suggestion clean ur pc it might be dust thats blocking the heat in
Open test bench atm.
So far, the list of solutions that failed me.
- Changing cables
- Lowering Refresh Rate
- Changing Graphics cards
- Cleaning the PC
- Replacing thermal paste
- Ramping GPU Fans to MAX
- Frame Limiter
maybe just buy a new one i dont think u can use a potato anymore
Bruh, i ran any other high end games, its just fine. So, until the devs themselves message me a solution, and the fact that i just used this GPU on my sister's computer and runs just fine, not yet.
monitor frequency will not affect card load unless vsync is on
Maybe try to change ur graphics
@eager shard
Forgot to add, everything is LOW.
Hmmm did u try del game and install gain
Yes.
Did u check task manager
btw you GPU should be fine at 80C and 100% util so you dont have to worry about the card getting damaged
yeah. it's been said before, many cards nowadays are designed to run normally at 80, 85c, a couple even 95c (with a sprinkle of doubt)
@dusty trench what framerates are you getting without vsync or framelimiter?
200+
then theres no need to worry
set a framelimiter to 50 and check that it's working
Im not worried, I JUST WANT TO PLAY THE BLOODY GAME, and i already tried the framelimiter.
if you get the same temperature with and without framelimiter, the framelimiter ain't working.
I used RTSS, set the game to around 40, and it still gives me lost signal, this lost signal thing happens after the update.
Around 5-6 days after the update. I was grinding Fallows.
lost signal?
as in the card fails to output video?
Yes, GPU lost signal to monitor, i have asked to my computer vendor, they tested the game with their monitor and own graphics card (3060 in this case, they told me) and they also managed to get the same problem.
So i am really confused atm what to do now.
the issue is self evident, why don't you rma?
Its been weeks, and i am bored, ordering a new graphics takes time and money.
um... RMA is supposed to be called precisely so that you don't buy another one
I dont think i can, its been a while with this graphics card, and RMA process is going to be a pain in the arse.
if you've verified that it's neither cable nor monitor then it's worth a try. I'll keep you entertained too.
I tried both monitor and cables, in fact, the vendor guy plugged into the repair monitor to check if my pc can run the game, and the game loses signal on their monitor too.
Also, after the trip the shop, i bought 2 HDMI cables, tried both, same results. So, that suggestion has been tried. You may refer the list of things i have tried above. So until support can answer, i can only hope, or i have to buy a new SSD, a new OS and the game in that SSD just so i can play that game.
Which is very irritating.
oh sorry, I meant that it's worth a try sending the card to RMA if you've already crossed out monitor and cable
(because then it's very evidently the card's fault and you're less likely to get rejected)
Its alright. I guess its time to get a new SSD, and see if the GPU is the actual standalone problem.
Yeah my asus 3060ti even with a slight overclock I did myself doesn’t go above 65c (with a custom fan curve) so that seems odd to me. It’s usually at 59c in cotw though. I’m also not running it on my ssd because it’s not a competitive game. Something seems up with your gpu.
Bois. I finally found the problem. One of the files in the drive caused the game to do it. As i put the file in a single drive, it works! No lost signal!
Others have said this but pretty much all Nvidia cards are fine under 90, many of them peak load at 80-82 C.
Yeah my gigabyte 3060Ti does the same for some reason. What settings do you have for COTW?
I run everything on ultra except for shadows on medium
hmmm i have everything on ultra too but i get up to 85
There's also question of ambient temp VS GPU temp. If your room in 30°C than GPU will throttle higher no matter what...
this might be the PC case + air-cooled CPU combo with poor ventilation...? Have you tried monitoring your mobo temps? If it sits below 60something°C then you're good.
mobo temps cant really be tracked but if you meant cpu temps ill give it a look
max 70 at some points for like two threads
there is a sensor on mobo, at least one. You'll have CPU, GPU and then most likely something labelled as system, maybe southbridge.
try HWMonitor by CPUID if you want to see this under Win
that's really good under the load... 🤔
can i find that in msi afterburner?
oh.. I have no clue 😅 quick google revealed nothing but seems that there are tickboxes underneath this icon - maybe it will be there? @vivid oasis
i have every graph on already and cant find it :/
this is where HWMonitor will shine the most I reckon 😃
it might not be as pretty but you will see all the raw data that mobo can produce in real time
mine's got TOO MANY sensors 😆
80c is ok above is overheating
85 too but its doing it even on lower settings
so the mainboard is essentially slow-cooking... Now we need to figure out the culprit. How are nearby temps look like. Hard drives, rams, whatever you got
I mean - you're idling at 85°C or is that under the load?
under load
oh, that's pretty normal for clocked things. Mine wheezes under load too. Maybe not exactly 85° but 3× as hot as idling. What's the resting temp?
@rare gate uhhh (this is COTW IN THE MENU)
The power fluctuates I see, do you know if that is normal? Does it do this only with cotw? I was thinking mabe you have a capacitor swellen
ill check another game rq
Try with a lower spec game then with a higher one if you can
And if the results are the same, you are looking at a hardware problem in my oppinion
the power draw would change depending on how much it's calculating. So technically it should significantly drop the moment you kill everything heavily relying on GPU. I have a different card but will replicate what you're doing. this is the starting point:
- mordhau 2) escape simulator
ahh the opening music of CotW.... beauty.
so the power went up from 30W to 150 and jumping around all the time
we have to bear in mind that 30 series were never meant to be optimised - they were made for RAW POWER, especially Ti variants while 40 series is just newer and less power hungry.
this is in the menus of COTW
so whatever CUDA or Optix modules are used in CotW are revving your card HARD
same with Mordhau
to be honest though your hot spot is 10® higher than what internet claims your card to run at so I am kinda with @cerulean canyon on this one
what I don't like is how uneven your fans are - look at the speeds.. It might be how she's built but that's kinda unlikely...
I also agree, a lot of people have come out and said they had issues with gigabytes 3060Ti's... I've also already sent the shop where i've bought it an RMA so hopefully they can pinpoint the problem and maybe give me a new card. I still have my old one so i dont have to eagerly wait for my new one then :D
phew! that's a relief. There's nothing worse than waiting for new hardware..! 😆
i just realized that 😱 WTF are those speeds 😭
she sounds busted bud..
saying that I'm no expert. Maybe one of the fans is doing something else. Maybe they do not share the heat sink etc etc. All I'm saying it looks sus.
yeah im sadly no expert either but i have a few tech fanatic friends and one even works at a small tech repair company so theyre always a great help :D
Also thank you for your help! @rare gate
The workload from the CPU is being diverted into the GPU causing an overuse of resources and forcing the GPU to overexert itself forcing the fans to 100%
^ is what a friend of mine said
my pleasure dude! And it does make perfect sense. I never assumed the engine would be so sophisticated to push the load into a GPU from a CPU 🤔 I've never heard about such game.. But if they calculating the same thing it does make sense to use the one that can do MORE harder. Still it does not explain the disparity in fan speed and why is it reaching over 100®C
might not be the problem because i asked some other friends who are also tech savvy and they said it might be the thermal pads and paste because Gigabyte is known for their bad quality of said things. @dusty trench maybe this is the problem with yours too
Yeah, i think ill be ditching Gigabyte graphics cards for now
Ordered myself a 3090 from INNO3D, and possibly going for a waterloop setup later, but god, China shipping is quite long
Interestingly, now that i dont have that lost signal problem, my PC now wants to kill itself. As i jump into the game, it kills the PC.
do you have CPUID HWMonitor? maybe check the GPU stats and share them here so @rare gate can help
Ill go grab it
brb
when you say it kills the PC - what does it mean? BSoD? Instant shutdown? Any mobo beep? Complete freeze?
Yep, instant shutdown
no, sounds, no nothing? Just whole PC turns itself off?
yes
that sounds like either overheating or power supply problem. What is your PSU? How many watts, what tier?
850W Gold Cooler Master White
GPU at idle, no games playing, applied new thermal paste (Kryonaut)
and how does idling mobo temps looks like?
Z590 Vision G is the MOBO
rolling out with style I see! 😎
Those temps you just sent, is that idling?
Yeah, currently just watching YT and downloading Por-, i mean google photos
ahahahah 😆 fair enough!
MOS looks rather hot for a mostly idle state... 🤔 does your case blow an extra air in?
so you have an open case with components exposed to the ambient temp/room?
Yeah
that means that those fans are not effective at all.. It looks brilliant for the photos (and kudos for stylish build BTW) but the fans instead of pushing air in one direction they just disperse air around them. You need a tunnel...
Hmm, i always thought of the front fans just blowing air into the rad and thats pretty much about it, the bottom fans shoot the winds into the GPU
And in case ur wondering, the RAD is Suck and blow, dual fan setup
long story short - put the panel back on. Also.. you know that hot part in VRM MOS? That's what is getting hot due to lack of cooling.
Aight, no probs.
this should illustrate it better - If I am right you're pulling air from GPU back into the case - unless it's reversed?
The fans blow into the GPU, there might be slight winds from the fans that will blow to the case.
So, you want the back fan to insert air to the case?
even if it is not how I drew it, if you put the case panel back on you should notice the mobo temp drop.
the configuration that is right is this:
and I am assuming this is what you have?
Uh, if you want the botton fans to blow towards the bottom instad of the top, then the winds is simply blowing wind towards solid metal.
Based on the recent pic, you are wrong on the bottom fan, its blowing wind into the GPU. Other than that, everything is right.
so other than the bottom (which is very arguable because case width, fan speed and case distance from the floor) pumping air onto the card it's ideal. THe only reason I always push the air down from the case is becasue I HATE dust in my case and it always settles underneath it. Closed case should take you down around 10°C on the mobo more or less
Gotcha
but that still does not explain why shutdown. Maybe run something at the background to see what happens just before the shutdown. To log the temps and Wattage before shutdown - if that makes sense. HW can't do that if I remember correctly
Would be nice to screenshot (not picture, but to log) the situation before the shutdown
AIDA64 Extreme does that and A LOT more but it's not free.. Maybe worth talking to Captain Jack Sparrow? 😉
I do have a cracked version of AIDA64 Extreme}
(my money to USD is a b'tch, so yeah, sorry AIDA
Can it toggle log the numbers? Im not very good with AIDA
this should answer your question.
It even does alerting(!)
Aight, lets give this a go
Ill try with GTA V first.
yours seems to be more complicated than standard issue - it's worth running full diagnostics
Ya want the full stats of everythin?
well I doubt the hard drives can casue an issue...
the things that would shut you down are things that causes mobo to crash.
You have a new mobo.. 🤔 thinking about it - maybe BIOS did save some logs from crash? I have no idea if your mobo has that.
But if you're running AIDA than what we definitely need to look at is things directly plugged into mobo
RAM
GPU
CPU
PSU (if we can - but that would be just Voltage log of the mobo)
Hmm
Aight, what can ya suggest that need alerting
Huh, suddenly it decides not to crash now
I capped the game at 60 fps, my pc is weird.
you'd have to go and learn what's too much for each component unless AIDA can do it for you.. Rather than that just keep the log running and as soon it crashes open it to see what seems much higher than idling PC to try and narrow it down. Than use that data to double check your findings online. Can't be more complicated than that 👍
Hello everyone, an update. Someone recommended me to undervolt the GPU to avoid this issue from happening since the GPU's voltage and clock was exceeding its limit from the game and power usage was very high. So undervolt has solved the issue.
(Its kind of embarrassing that EW hasnt contacted me in support for weeks now.)
Where did you change that in? I think i want to try this too just to see if it does anything for me too
MSI Afterburner
Oh yeah, update, it seems my GPU is actually faulty, the moment i stress test it, it dies. Welp, thats 300 bucks down the drain.
Damn that sucks. I think you can still send it to RMA to get a refund
Meh, its going to be too tedious, so ill hold for now. My INNO3D 3090 ICHILL X3 is in transit
I was tempted to buy a 4090, but eh.
if mine is faulty too i think im going to buy a 40 series too probably XD
Yeah, oh yeah one more thing. Never force shutdown or use the power box to shut down the PC, now i understood the problem. My mom keeps doing that because she lacks PC knowledge just to mess with me over some oversleep
whys that?
Shorts the GPU.
It can break any computer if you do, to be specific.
btw wdym with force shutdown or using the power box cus i just use the power down button in the start menu
What i mean is, flipping the switch of the PSU, or using the power box to literally knock the whole electricity of the house. That will damage computers
yeah okay that makes sense 🤣
Hello everyone, one final update. The PSU was replaced with a Corsair 1200W Plat PSU and immediately works. So, to everyone, please use a surge protector to protect your PSUs. My 850W is toasted as i do research. The Graphics card is fine, thank god.
Tested with Furmark, all good.
This is also a PSA
Try a driver rollback. Seems quite a few people started having temp issues with the new driver.
I am using an overclocked 1660 super with all settings at ultra and I am having no heating issues at all. Average is 43 c 60 max.
Hubby RMA'ed his 3660 and they replaced with a refurb. It has now died as well. He's now running a factory OC 1050TI with nary a problem. The 3660 seems to be junk.
If you replaced the thermal paste on the GPU, then RMA is probably pointless - the paste replacement likely voided any warrantee.
Yeah, i forgot about that, but now my RTX3060Ti is fine, just the PSU is blown as i figured out.
Don't get a cheap one. Do your research and get a good PSU. Try GamersNexus for reccomendations.
I don't think the Cooler Master 850w gold is cheap, quite expensive