#PC keeps Shutting down
888 messages · Page 1 of 1 (latest)
that does not sound like a software issue. That is far more likely to be a hardware issue.
it could definitely be the PSU but it could be another component
I've had similar issues when my ram was overclocked or cpu
is there a way i can check what component it is?
i havent changed anything at all
and it only happens in games that have my gpu at 100 percent and i have lsfg on non performance mode
whats your psu?
could it be getting overloaded?
yea it could
its a 650 watt i think
or 700
idk if thats low
i looked it up and it said its fine
if its 650 or 700 then it should be fine unless its one of those cheap power supplies
it probably is a cheap one
brand?
im still using the one i got 4 years ago
how would i check
it came with a cyberpower prebuilt
a good psu should last a while, far longer than 4 years
oh its a prebuilt
might have skimped out on the psu
worth a check
its a prebuild but the only things that are still the same is the motherboard and the psu
yea that should be good
What do you think it is
it could be many things at this point so i cant really help pinpoint it
Oh
cpu or ram related?
Possibly
if the process is reproducible, i would do things until it wouldnt occur again
None of my components were overheating
tbf it could also be a driver issue
maybe for the gpu?
why not reinstall it and try
I have had a lot of driver issues
amd right?
Yes
which gpu
Rx 6650 xt
i see
i used to have huge issues with my old radeon card
very similar to stuff you experienced
You switched to Nvidia?
You think I need to upgrade?
turns out my old card was faulty
6700xt it was
but id probably say driver issues are the first thing to look at for you rn
quick question , what power supply do you have? share model if posible
I check in the morning
Sorry man
All I know rn is that it's a EVGA
600-700 watts
But it never happens unless I have frame generation on
And that's only happend 2 times
Within 3 hours
It only happend on ark
And pal world
Ark ascended
Both games my GPU is basically maxed out
But something like GTA 5
I can run frame gen no shutdown
I'm either assuming it's the PSU not being able to keep up for some reason
Or the GPU completely failing because it doesn't have any room to work with
And the temps are normal too
gpu usage is not related to power draw
you can have the gpu used 60% and draw max power
no biggy
here is the best way to try
get hwmonitor
run yre game without lsfg
check the max power draw
then enable lsfg and check it again
or if you have afterburner , just use osd
altho im preety sure yre power supply cannot handle the load put by gpu+cpu
You might be right
thats why it shuts down , to prevent yre components to get fryed
or protect the power supply
You think I should just get a higher watt PSU to be safe
i dont think so , all you need is reduce the load
How would I go about doing that
il find a way and let you know , usualy msi afterburner can help you there
by setting a lower power limmit
GPU usualy draws more power then CPU
unless its intel's newest cpu lineup
those beasts can draw 300w easy
thats a good cpu ,it shoudnt draw more then 60w
so you should be ~300w
max
altho i dont know how accurate google is
seems yre not alone
Hmm
👉Need a repair? https://krisfix.de/en/geraet-einsenden
👉KrisFix Shop: https://www.gpufix.de
👉Question about repair: [email protected]
👉Need Help? Book your Technical support here: [email protected]
👉 Leave a tip for us if we've helped you out:
-PayPal : https://www.paypal.com/donate/?hosted_button_id=WN4CET3V79S9Y
-ETH: 0x03FFa9DF33F4edcE958a...
from the video , it seems more a power delivery problem
so limmiting it should be a quick fix
I could do that through afterburner?
sure
set it 80 for start
if it fixes it , il tell you how to set it default so you don't need to keep setting it back on boot
Ok
Will you help me tomorrow?
Like 8 hours idk if you will be on lol
I'll do all of the stuff
And I'll let you know
il try but in like 2 hours i go to work
its 12:19 PM here
and im working from 3PM til 11PM
Same here! My onexgpu and legion go is shutting down during gameplay with the 2.8 update. Before that it run perfectly!
do you use performance mode of lsfg?
Not yet. Should I? Guess there’s a problem with FPS cap. If I don’t cap it seems to work without crashing. The former version had not this problem.
@upbeat eagle how is it?
What do you mean exactly? I’ve restored the fps caps in riva tuner and ingame and tuned it back on. For now it’s working well.
If performance mode or not wasn’t my problem. My problem was that my LeGo and my ONEXGPU turned off during playing dragons dogma. That happened only with the new Lossless app. However, I’ve turned on and off the fps cap (30 fps) and now it seems ok now. Runs very well for the moment, even in performance mode. I hope that will last. ☺️
Dude, I have the same problem with the latest update of losless, my pc shuts down when I play something. I think it's something from the last update
I went back to the old update and everything is perfect
I thought I was the only one 😭
I am also used to do demanding overclocks to my pc and I have never had problems with the power supply. Something this latest update does that causes the pc to shut down.
what is your gpu?
But it would usually happen like 3-7 mins into playing then boom instantly my PC goes out
Rx 6650 xt
yes, mee to
8 gb vram
mi gpu is amd 7600
maybe it is some incompatibility, I hope the developers read this post.
is possible
post dxdiag here
also, if the machines shuts down, the issue is major hardware or software failure
that is not on LS
but we can try investigating
yes
yes
also have this issue
judging that both you you are with AMD and experiencing the issue, there was an issue with some of the newer drivers
downgrade 2 versions back
A few minutes after playing a game with the latest update, the pc shuts down and if you turn it on it's not like it shut down unexpectedly for windows. It is like a normal shutdown
ok yeah mine dif than his
and remove the LS profile from there, reboot the machine and try playing again
in that case downgrade
ok, I will also check the windows event viewer to see if I can draw any conclusions that might help.
Thank you
if there is a crash log- it might shed a light
but usually general system crashes show BSOD and the problematic component down left corner
Do you have the problem too? @errant pivot
yep
Can you pass your dxdiag report please?
yeah mine dont do a normal windows shutdown my whole pc just goes off
It may be some kind of incompatibility with the latest drivers as ogioto said before.
yeah
Can you test if using the performance mode that they have put the same thing happens to you?
videocardz has links with mirrors for older amd driver versions
with performance mod zero problems
that means your GPU is weak for the quality mode
and yeah, 780m is not too strong iGPU
it happens sometimes but sometimes it does i really do think its a driver issue
doesnt*
But my graphics card is equivalent to a 3060, I'm not sure if that's the problem.
that was for the iGPU of KASUAL, for your try the driver downgrade
before downgrading- uninstall the current drivers with DDU
that will clear all leftovers from them
should i downgrade to 23.11.1? im on 24.3.1
mew
hey
i have 2 gpu
you have a more powerful desktop graphics card, try running a game in performance mode to see if the pc shuts down. @still depot
before reverting the drivers I mean
on laptops LS usually defaults to the iGPU
you can try switching that
ive been using quality mode for the past 10 mins now and no crashes
If you have not changed anything in the configuration of the PC, it is normal that eventually it will happen again.
im trying to get the crash on video
well there are some stuff i have changed
but i dont messs with gpu or cpu oc
or configuring either of them
if it is a drivers issue, it will be a bit random, not fixed as when it happens
yep
Anyway, thank you for the work you do, I think it's wonderful! @plush falcon
try 24.1.1
Ok I'll downgrade
It also only seems to happen when I use an app called audio replay
Relay*
Because it didn't crash at all until I used it
To send audio to my phone
Ok I downgraded
I will do the test again
give me about 20 mins
ok
to see if my pc turns off
Yup this time it was faster
Within 5 mins
Im gonna try performance
Now
i am also using the scaling
idk if that means anything
I only use it for frame generation and it also shut down.
If it still happens with the driver change, it is even stranger. Maybe the quality mode application is doing something it shouldn't be doing.
Just to clarify my PC doesn't just shut off the screen goes black for about a quarter of a second then shuts off
put stats one on the display and watch the temps and so on
yeah, watch for unsuall stuff before the shutdown
watch the utilization, temps, etc
and check if some instability happens before the shutdown
You have the same cpu as me, can it have anything to do with it?
im gonna try to record the shutdown
idk
you said yours is a windows shutdown right?
my whole pc shuts down, but I'm not sure if the monitor goes black before or it shuts down directly
Then I turn it on and there is no typical windows message notifying of an unexpected shutdown, everything is normal.
when i turn my pc back on
chrome and wallpaper engine acknowledge that my pc unexpectedly shut down
Did you try the performance mode and did it fail as well?
nope
worked fine
rn quality is working fine
its always when i try to catch it happening too
Then with the performance mode without scaling. Only with frame generation it does not shut down?
when does yours usually shut down?
i havent tested to much yet
I have only tried with the standard configuration, with lsfg enabled and performance mode disabled and it would shut down. I have not tested with the performance mode
oh
mine will run fine with all the settings
but after a while it will shut down
perf off and on
with and without scaling
but within 5-20 mins my pc will be off
I'm going to try red dead 2 with the performance mode, now I'll say something.
set custom resolution 1.1 and enable window mode
best way to understand what it does
use windowed non-borderless in rdr2 to not get strange yellow borders
also, what psu do you have mew?
in theory with the new update and the performance mode in the lsfg how much better does it perform compared to the previous update?
performance preset uses @3% more then the old one , but it looks way better
Does the quality mode use much more?
yes
do you have stronger PSU?
doesnt have a grill under it?
might be power related issue
or problematic psu
the new model is heavier and hungrier for resources
also, that way the GPU could be sending all the hot air into the PSU
if it is sending the hot air into the PSU it might be cooking the capacitors over their limits
thats a strange one
I would recommend to borrow stronger PSU and try with it
I can just get a new one
From best buy
Probably shouldn't have bought a cyber power PC 4 years ago
neah
i have a 10 year one , works fine
but you should check the temps with hwmonitor
in 4 years , cheap thermal paste may dry
can you check temps under load?
it has a free version the right one
yes
if you get new psu, get a strnger one
hotspot 101
yeah
you reach the tjunction point
the system shuts down to prevent damage from overheating
also, download the normal non-pro hwmonitor
to see all stats
Is there a fix?
yeah, clean/change thermal paste or upgrade cooling
first one is recommended
if the issue is caused by small case tthat might not fix it tho
you might have to check your case airflow
i dont know , a simple repaste should do the trick , only if the problem persists , then i whould do that
Repaste the CPU?
didn't you said you repaste it 6 months ago?
new?
yes
brand new
but i do still have the cyberpwoer case
not my pic
but the airflow doesnt look to be great
there is a grill at the top
and back
should i see if undervolting the gpu will help?
-10 for start
it only let me goto -6
thats a strange gpu
go with that
btw does this version of HWMONITOR work for you?
you need to run the x64
and to close the other hwmonitor that you have now
i dont think the x86 and 64 make a diference in this case
win11 dropped support for 32 bits on lots of stuff, so it does actually matter
yeah, the GPU is overheating
being too close to the PSU probably cannot push the hot air out
wait
it wasn't the GPU on one of the pics
no
so i have to reapply thermal?
it hits tjunction probably, and maybe upgrade the thermal pads
yes , but get like a arctic mx4 or 5
50 degrees on idle isn't kinda much?
its idling at 100 on ark ascended
nah, my idles between 45-50 as well
the issue is the high point
usually the memorry is the hottest component on the new cards due to high GDDR6 temps
what about 6
so definitely upgrade the pads
check the specs for yours
and get variants with higher conductivity
because the pads cool down the memory
but can you try running the HWMmonitor I sent earlier?
it can give mroe stats without the "TRIAL" thingy
yeah its running
and shows all without TRIAL, right?
yes
oh ok
so i just need to reapply thermal paste and apply pads?
and maybe undervolt?
ok , im confused
first you said you didnt know you need thermal paste on gpu
now you talk about pads
oh thats because ogioto was telling me i need to
i see
i didnt know any of this before today
okies
most of the time , the pads are ok
the thing is , when you ran LS on quality mode , yre gpu maxed out
and the way yre gpu is proteced himself was just shuting down
nah relo, totally normal stats for when you put pressure on the gpu
I mean yours
I'll show mine in a bit because I update hwmonitor
And I said about pads
because I will demonstrate in a bit
the new cards use GDDR6 which is the hottest element it them, it can go beyond the GPU core
my card has more temp sensors
so wait for me to destroy a building in the finals and I'll show you both the stats
lol okies
the gtx 1060 is fine from meory side due to gddr5 running cool
all the new GPUs suffer from the heat of gddr6
here
that is the modern plague
i guess the good old days were better
and the new gddr6 memore is not colled as the old cards with paste
paste is for the core
pads for the memory chips
so both need upgrades
it depends , they have diferent thickness
any works but there are 2 factors
but i still think a repaste should do the trick
I ordered artic MX 4
1st- you need to get pads with the same thickness as your current ones, stacking pads doesn work (if you need 2mm for some chip stacking 2 1mm on top of each other will cause a problem)
order? you don't haev a store close? you pay more for transport
2nd- you need to get pads with higher conductivity than your current ones
Does best buy have them
Ok
again ,im unaware how things work in america ,or canada
I'm scared to open my GPU tho lol
you can check online for info on the pads for your model, different manufacturers use different ones
dont be , plenty of guides on youtube
if you cannot find online- ping the manufacturer support
Ok
@plush falcon no sensors for memory temp here
keep in mind that if you only change the paste and need find out there is still high temps going, you will have to repaste it probably again when the pads arrive
gddr5 didn't had sensors for that
only the rx7000 and newer amd cards and rtx 3000 have such
makes sense
before on rx5000/6000 and rtx2000 Jayz and the others were using additional hardware tools to measure them
but gddr6 currently runs higher temps than the core to be able to reach such high transfer rates
I tested it on more cards as well, always the hottest component
i get it , but still , a simple repaste should do it
no need to overcomplicate
especialy if the pads are fine
i know , you like doing all of them , just to make sure everything works
I'll get pads just in case
okies
If the repaste isn't enough
hes not wrong tho , it could be the memory
but most of the time ,paste does the trick
the pads wear out like the paste
the higher the temps- the shorted their life becomes and they wear out
im fully aware , but the gpu has 6 months
possible
but for example- EVGA recalled lots of rtx 3080
because they dopleyed shitty pads
mine is one of them
I decided that I do not want to send it to Germany
clearly
an decided that I will change them after a while myself
because didn't want to stay 1.5 months without GPU
and my friend kept comfy 70-75 while playing
while I used LS to keep it going above 85 xD
so, yeah, some manufacturers still use shitty pads
especialy OEM
this is my gpu curve
the stresstest was without
i dont need more then 1.7ghz
and thats 350mv in my case
to go to 1933
and trust me , 350mv is alooot
same stress test
0.750v locked
-7 fps lost
so yes, undervolting does help , but also underclocking
worst case scenario , sistem reboots
undervolting doesn't damage hardware , only overvolting
so this can be a quick fix for yre gpu ( dont use my settings , thats for my gpu , yres has other )
no biggy , sorry , but you are kinda limmited
Any news on the problem?
yes , it usualy shows issues with gpu's
overheating is one of them
framegen is gpu intensive , especialy useing the quality presset
but people are too stubern to switch to performance
so when a high gpu load happens , and the device cannot handle it
it just shuts down , to prevend hardware damage
or BSOD
illl stick to performance
and apply that thermal paste you recoomended me tomorrow
and if it still overheat ill do the pads
go with thermal paste first
trust me , its a easy fix
just make sure you don't loose screws
lmao that would be really bad
indeed
i know its sounds like a hustle , but its realy not
its just liek sex , only the first time hurts
after that , only pleasure
😭😭
i know , not the best comparision , but you get my point
Mew, what temperatures with lossless do you have maximum on the GPU? @still depot
in Quality mode
on ark ascended the hot spot is 100 celcius
and lossless scaling pushes it over 110
and my pc shuts off
with performance mode idk
nope really
my pc has not turned off at all with performance
wrong chat :))
thats so stupid tho why would amd push cards out that are almost at the max
when under load
just because the manufacturar sais that 110 degree is fine , that doesn't mean you should
they wanna sell gpu's
sometimes i wish i had gotten a 3060
ofcorse they whould say something liek this
tbh , 3060 is realy good , especialy the 12 gb one
but 4060 , waste of money
i dont know , you shoudn't give up on yre gpu
also did performance mode helped you?
20% gpu usage is substantial
bricking yre device for quality interpolation is not worth it
sticking around 100 c with performance mode
its the same without lsfg on
but playing ark im gettin 97 percent gpu
the game has lumen enabled
even on low settings
yeah with performence i stick to around 100 c
ohhh
yeah with quality
it spiked up to 108
not good
indeed
30 usage percent in quality mode and 65 temp maximum gpu
is weird
30 percent usage only lossless
in kingdom come deliverance i can use quality mode with any problems but in rd2 is not the same
wth
better question ,why is yres overclocked?
mines overclocked??
see the lower line on the fist graph ,thats the default
wait im confused
hit this button , check if the graph goes back
that left one is from the video
i was just wondering why his looked like that and mine looks like this
i dotn know , its just it is overclocked
this one is the one you need to focus
doubletap somewhere to remove the highlight
noo thats in the video thats not mine
what is yres?
yes
yes
gimmie a second , im still unaware how amd voltage controll works
again ,im unaware how it works
i see how it works on afterburner
the top one controls frequency , the bottom one the voltage
the bottom one is what yre intrested in
the point on the right , drag it down 100
holy
share yre screen afterwards
its definetly somethign wrong , the difference between the two shoudn't be this high
that would have been cool tho
just drag the two 100 less
from right side
share yre screen afterwards
that means you lose 100 mhz for 100mv
only lets me go down 75
if the frequency cannot keep up with voltage , worst case you get a crash
put that back
ok
the right side of the floppy
now run something and check the temps
this is mostly a trial of error
most of the time -200mhz is a good start
then reduce the voltage as much as posible
as long as you do - , eevrything is fine , even if it crashes
ok
ok
set 3d profile to 1
ok
hit apply
ok
go back to general
ok
ok
this way you dont need to re apply yre profile
and if you wanna try to undervolt it even lower
just save it on 2
and if it crashes , it will autoload 1
but if -100mz was ok , you could try both -200
after that , only lower the bottom one
tbh , i kinda like that way of undervolt over nvidia's bs
there are a few forums where you can see directly people that experimented with the same model and what was the lowest stable that you could achieve btw
valid point
easiest solutions from people who spent hours with different configs
-200 did crash my drivers
try -150?
checked one forum
people recommend 1100 for 2.7 GHz
it can be dropped a bit more
but they said that was the last solid point of proved stability for sure
will check another place
ths runs 750mv on a 3090 , it all depends on yre workload
yeah but he runs that for compiling
exactly
he compiles models that way
games push more resources
compiling pushes memory mostly
a -200mhz wont hurt yre performance that much
it is also a question of stability
also valid
thats why i recommended to do both - the same value
I saw recommended 1100 for 2.7GHz and 1120 for 2.41Ghz
the second variant will hit the final fps
a little to big the differece
but should keep cooler
yeah, normal for undervolting
noice
play some games
for like 10 min
if it won't crash , then you could save it on 1
but only then
1 is yre backup
2 is yre test
btw, after one longer the finals session earlier-
that is why I said the new GDDR6 plagues are the temps
80 degrees , jeeez
keep in mind that it is the cold season and I do not use RT
otherwise I hit the tjunction easily
😄
well i have a room fan for summer in case for that
usually at the hottest 2.5 months of the year I can reach 93-95
my fan is next to the pc whole summer to not go beyond the mentioned above
xD
very nice 😄
yeah, fan and better paste are the solutions for you
my model came by default with 12 conductivity pads where it needs 15 to work in better temps, I ordered 24 conductivity xD
drivers crashed at -150
decided to go all out, my plate is very sturdy and it takes the heat
okies
try -200 on top -150 bottom
150mv is preety much already
didnt work
okies
remain on 100 for a while
im still not 100% sure how AMD works
Today's video we look at how to undervolt your AMD RX 6700 XT, 6800 XT, 6800, 6900 XT and 6600 XT with some simple settings and easy way to test it!
Sponsor Get a Windows 10/11 Pro Key for under $15 (Use 30% Coupon BFTYC) - https://bit.ly/BFTYC
Windows11(21$):https://bit.ly/BFTYC11
Office2021(49$):https://bit.ly/BFTYC2021
Application used, uni...
im lagging while watching a youtube video with -100 on each lol
tech yes guy is preety good at this stuff
with LSFG?
strange
im off to sleep
if you use amd software to undervolt , make sure you disable the settings applyed to Afterburner
or just straight uninstall it
cheers
Before that you need to default the settings or the afterburner settings might remain applied
default them via afterburner and then play with amd tools
getting around 87 celcius
Check if you have set a fps lock somewhere. I had set it to max 60 in riva tuner and that seemed to cause the problem. After I had removed the fps lock and reset it again it worked without crashing anymore.
FPS lock is a must as i see it , especialy since dumb game developers think running loadingscreens on 1000 fps is ok
also i do have a 240 hz monitor and ive read that a bad dp cable can make your temps go up to 20 c
good idea
so you didnt lock FPS?
no
but the game that had me at 110 c i was only getting 30 fps
lol
so far the highest i have gotten since i undervolted was a 93
so i guess thats good
undervolting is good, framecap is also good
I have this exact same problem, since the update if i use lossless scaling the pc turns off
You could of just set a temp limit, most people do that now days for these issues
Will never go above 80 and downclocks by itself
Did you manage to fix this? I also have an amd card and lossless makes my PC turn off, i'm sure that my Power supply is working
The undervolting and everything was unnecessary, you're just gutting your performance . Just run a temperature limit and let MSI downclock and down volt when it needs to so you can keep performance.
The average GPU temperature for gaming should remain between 65°C and 75°C and never exceed 85°C
Indeed i should said that
undervolting just overcomplicate it
but the ideea is the same , the gpu is overheating and you need to cool it down
use pefromance on LSFG
it will reduce the load on yre gpu
i will try ty
yeah so
uh
my fans were not being used to their fullest
they were only being used 20 percent
Yeah, usually they stay at low levels of usage until the are needed, then the profile should automatically kick them to higher level
That is completely incorrect and is massive issue. If his fans are only at 20% when reaching high temps it will indeed cause problems and cause his GPU to overheat, throttle and shutdown the PC to protect itself.
Your Fan speed options should look exactly like this. If not , someone set a custom fan profile or some malfunction is occurring
I would go into a game and check if the fan speeds are increasing as your temps go up.
If they aren't increasing in speed to combat the heat , click the cog wheel on the far left of MSI afterburner, select the fan tab and tick the box for " Enable user defined software automatic fan control" hit apply then ok. That will import a MSI afterburner default fan profile.
Outside of the fact that all modern hardware automatically throttles
And fan curves are nowhere near linear with gains in cooling performance
What? There's a literal internet load of evidence against what you're saying here, it's also common sense that they do indeed do what they do. Fan speed up = Temp down. Entire branches of technology exists proving you wrong. Your logic is entirely flawed.
You're grossly misinterpreting what I'm saying
You said fan curves that scale with temp do not have a linear effect on cooling performance. You are wrong. I can run a benchmark right now with a different profile showing linear temps drops with the fan speed increases. As long as the thermal paste is transferring the heat to the heatsink the heat will be blown away.
Looks so linear to me!
Again, I noted throttling originally
What matters is real world performance
You know what linear means right?
You just linked me a benchmark from Blender for a CPU showing how loud a fan goes the higher the clock is.
Are we even talking the same language what are you doing lol.
The point being that the difference between low and high fan speeds isn't the difference between your pc shutting down lmao
Of coarse it is. Higher fan speeds would cool the GPU and not shutdown the PC.
How do you not know that ? This is all common sense
Waste of time
Don't say waste of time, I'm generally curious on your logic, it baffles me someone thinks that way
Explain lol
what is happening here : ))
It's all good I respect they don't wanna talk about it. Hope OP finds a solution. Peace
even at 80 c and abovei t would only be anround 50-70
i changed it to when it hits 75 it goes too 100
and i havent had any problems
because having the fans at 100 keeps it like 15 c lower
which is insane
10-20 or sum like that
so it wont shut down
because now i dont go over 80
at all
80c for what