#Shutting Off completely
225 messages · Page 1 of 1 (latest)
more likely an issue with the power supply causing the system to shut off completely
could be bad incoming power as well
voltage being off normal
like below 115v
do you have a ups in line you plug the pc into or do you plug the pc directly into a wall?
I think the cpu temps are good it's sitting at 42 C° on normal operation. I think it might be the power then....how do I fix that?
Also I plug everything into a power strip
And then that goes into the wall
GPU not CPU
and it's more likely a power supply problem or your power circuit you are plugged into
how many devices are being powered from the same power strip including your pc
and anything else going to the same wall socket/outlet?
the issue has to be either the gpu or the full system causing the psu to trip it's protections
my money is on the psu being the issue since it looks like it's over 10 years old
or you're using a daisy chain cable, although that shouldnt be an issue since it's a single rail design. if you have 3 separate pcie cables, use 3 and try it again
if that doesnt help, the only other thing ic an really suggest is buying a new psu and trying that. return it if nothing changes, but not until all testing is done, hopefully within the return window
I have my monitor, the psu and some outlet sockets for my standing desk....oh and my standing deak
if you have 3 cables for pcie, start by swapping out a single daisy chain for a third cable. if you cant do that, then you would move on to swapping power supplies
@rough glenhow old would you say the psu is? when did you get it?
the pre 2022 models are from 2014
I bought it in 2018-19?
new or used?
does it say on it the revision date? or revision number for atx 12v?
I don't know what that is...
does it have a manufacturer date on the psu?
im trying to find where that might be on the unit
oh! is it black white and blue on the labels or just black and white/gray?
does it say anywhere in the listing the atx12v revision
it owuld be something like atx12v 2.4 or 2.51
might also be in the manual if you still have it
im just trying to figure out if the psu is having a problem or if the gpu is the problem
it could be either
I would have to look for it, i think I still have it
ehhhhh
probably wont matter
so the 7900xtx is new or was it used?
refurbished?
It's new
i guess right now the best you can do is swap to a third daisychain cable
probably wont change how it works though
Could it be some setting wrong with the psu? Like occ? Or something like that?
well either the psu itself has a problem and it tripping it's over current protection
or the gpu is causing it to trip it's over current protection with transient spikes
normally transient spikes wouldnt be an issue with a single rail design psu
assuming you have sufficient current on the 12v+ rail
with 1000watt psu you have nearly 1000watt on the 12v+
I checked before on CORSAIR LINK windows and it said 12v+
yeah that is just listing the wattage on the 12v rail
Elaborate?
your psu comes with 3 cables that have 2 6+2 pin pcie (usually for GPUs) connectors each
usually called a daisy chain cable
instead of a single cvable with a single 6+2 pin connector on the end, it has 2 in line (the cable) for daisy chain cables
Ah yes I do, got 3 of them to thr gpu
use all 3. just plug in the end connector
Oh already is like that
ok then i'd say it's probably a gpu issue
the psu could be tripping, which likely means a problem with the gpu if it does a full shut down
it could also be the gpu is overheating
are the fans very loud when playing a game?
To be honest I don't know, I play with headset....oh wait I noticed a vga light
On the mobo
return the gpu and try to get a replacement or try to get a 9070 xt
But isn't the 9070 xt worst?
it's like maybe 5% slower on average but has better ray tracing and fsr 4
oh and less vram
How is it for 4k gaming?
unless you were playing at 4k, losing 8GB vram down ot 16 is not a big deal
well you would likely be using upscaling at 4k anyway so 16GB probably isnt gonna be a big deal
depends on the settings you were using, but the 4080/5080 were also facing the same issue of possibly not enoguh vram for 4k and most people dont have much issue
To be honest I bought that to hold me off till the rtx 60s family
i've tried 4k on my 3090 and rarely saw it go above 16GB with some exceptions of like 19GB but that was allocated and not actually in use
Planning to get a psu more powerful by then too
if you are still in the return window for the 7900xtx, yoiu could probably get some money back and then try getting a $600-700 9070 xt model
if the 7900xtx is defective, you dont really have much choice
better to return it and get a replacement
and if you cant get a replacement and just a refund, get a 9070 xt instead
Sounds good thanks, so then it's not the psu? Can I test it somehow just to make sure?
well cant be sure but it could be the psu
you really kind of just need hardware to test with
swapping gpus would be easier than gutting the pc to swap in a new psu
i mena probably not by much but it's still easier
and more likely the gpu is the problem
if you had the money, you could go buy a new gpu and psu and return the one that isnt the issue
it can still have 8% or so degredation over time
it cant do that huge wattage smoothly anymore
id check gpu temps while playing for now
open adrenaline and turn on the monitoring
the efficiency drop over 6-7 years wont be huge
and the system itself will only be using at most probably about 600watt
the psu likely would handle the transient spikes just fine
i'd like to know what the gpu temperatures are and fan speed, because that could indicate if the GPU is overheating
if it were, that would still mean needing to replace the gpu, get a refund or RMA if those options are not available anymore
you could do a repaste yourself but you risk voiding the warranty if you damage something
370watt tdp on that card only slightly higher than my 3090 and with possible oc headroom that could hit upwards of 425watt. total system wattage before transient spikes is probably around 590-610 at full load or typical gaming load more like 560-580watt
gpu oc for gaming would probably be more like 600 like i stated before
maybe lose up to 12% efficiency over 6 years leaves around 877watt
Between 46-57 C°
do you know the fan rpm?
high fan rpm with lowish temps when gaming, can sometimes indicate a beginning overheating issue, but im kind of doubting the gpu is over heating
@rough glenyou need to make a decision on what to try replacing first. the psu could be at fault and the gpu could too
it says in the amd control panel
I'm not at pc right now, will get back to you on that in 2 hours or so
How fast is corsair with warranty claims?
no idea. i havent had a warranty claim with them in forever
last time was when i was rmaing some ddr2 ram that they gave me the full retail price for even though i used a rebate so they actually paid me to take back defective ram 
but yeah maybe rma'ing psu first is a better move. you just wont have a pc for awhile if you cant buy a new psu
Because it's supposed to be 10 years no?
without knowing when it was manufactured, it could be up to 10 years old
or it could be 1 year or less from when you bought it'
2015 seems to be when they started making them
right now im using a corsair ax 760 760watt i got in 2015 but was manufactured in around 2014 i think and it's a 2013 model
i saw reviews going back to 2014
op said pre 2022 models
saw a toms hardware
yeha i saw the toms as well
it's a good psu but the efficiency may have dipped too low for it to handle the transient spikes anymore
yeah if it is 11 years old
which is what i've been suspecting is the case with my ax 760 being from 2014 and my 3090 crashing at 100% power but not at 90% power limit
warranty is calcuated to be 10 years for a reason
@rough glenthere is another thing you can do. you can lower the gpu core clock and voltage in the amd control panel a bit and see if it still crashes or shuts off the entire pc when gaming. if it doesnt, then it could still be the gpu, but may actually be the psu
check temps first though
my warranty was 7 years on the 760watt so it ended in 2022
i really need to get my rm850x back in my case
looking at what i've said so far and it's getting a bit jumbled
if you need more clarification on what to do, let me know
I'll check my temps soon enough, is it ok to mention you?
for your tech support post yes
wlel anywhere i guess
Oh uhhh, you probably need gpu-z to check temperature while gaming
gpu z?
It can log sensor data and shows a graph you can check the highs and lows as well as a reason for some cap on performance
this?
what am i supposed to look at?
Sensors tab
you need to actually be playing a game, but it also looks like it's not properly checking sensors
you might have a driver problem that could possibly also explain crashing
should i used ddu n do a fresh drivers install?
no
shut off gpu-z and start it again
you may need to close it from the task bar near clock
still the same
should i reinstall drivers?
nah
i guess amd cards dont work properly with gpu-z reporting on sensors
i dont even remember if i used gpu-z with my 6800 xt
i mainly used the amd control panel to control settings and msi after burner for in game overlay reporting
i'd just uninstall gpu-z and delete it since it wont be of much use for you
so then my best bet is either get a new psu or gpu then?
well there are some tests you can do like i've mentioned
in the amd control panel where you control the core clock and voltage, you can lower them to see if fixes the system crashing or going fully turned off
but if it's only a full shutdown happening, there isnt any real way for me to know if it's gpu or psu
can just use the amd overlay
i forgot that was a thing since i never used it
my brain is screaming at me to go to bed
???
what is said yesterday, easiest way to do it
alternatively go here and hit it
or open a game and presss f11 i belive
So I'm looking for gpu temps right?
It's 62 C°
max in-game?
Should I overclock it for the game? I'm just running rivals
you are having power issues
Damn
thats the whole point of this post
Oh it just turn off like right now
its supposed to restart
not sure what you did
if you pressed the button it does restart
but you are most likely in need of a psu
The screen blinked then the fan for the aio turn off, the psu fan spinned then it all died
psu overheat/other internal issue
if its less than 10 years old
well have fun
I was saying lower the core clock and voltage which means underclocking. How did you get overclocking from that?
I wanted to know if underclocking the gpu would stop the system shutting down at all
No I know, I did that myself. Also I underclock it, but haven't had any trouble. It just goes off after a while of usage
So... no trouble gaming with an underclock but you say it just goes off after long usage? Is that any time whether playing or idle?
100% of the times it happend was playing
And also I haven't gamed that much while the gpu being underclocked. So imma have to try that today....
Once I think
It shut down while gaming 6 times already
Well, I guess try rma on psu and then if it still happens it's probably the gpu
That's what I'm gonna do then!
