trying to help my friend

[MightyNerdy93]

Hey guys , didn't where exactly to open this thread so if its not in the right place , i would like to move it please.

so my friend has this rig:
14700k
z790 max
4070ti

recently when my friend plays with me fortnite , the computer just shut down (itsn't BSOD)
so he called the store where he got this pc from and they told him that maybe he needs to change the settings on the menu of power options on the OS and maybe there is a need to replace the power button on the case (4000d)
after 2 days , now it happens again.
what could be the problem ?

 

[Lost_Troll]

It could be a bad power supply or the something is overheating.

I would have download HW64 info and run in it log mode while they are gamming. They can pull the log file and see if the power drops out or if the temps get too high.

 

[QuietBob]

Might be a power delivery problem. A stock 14700K consumes 280 W fully loaded, and a 4070 Ti may spike to 320-350 W depending on the model. I wouldn't run such a config without a quality 750 W unit at minimum.

Unfortunately, pre-builts often come with cheap power supplies. Could you get the PSU model from your friend?

 

[Chrispy_]

Get your friend to install monitoring software like HWInfo or OCCT, make sure to output monitored results to a file.

Then, after the inevitable shutdown, check that file for temperatures, power consumption, voltages, etc. If there's nothing in there, get them to filter the windows system eventlog for the last 24h by criticals and errors.

14900K is a power-hungry, hot-running CPU, and the 4070Ti isn't exactly sipping power either. As hinted at by previous replies it's likely to be an overheat or inadequate power.

I would be using an 850W PSU with an 80+Gold rating at a bare minimum, since that combo of CPU+GPU could, potentially, pull about 1000W for a brief instant and anything other than a relatively decent high-Wattage PSU is likely to struggle. The claimed specs of the 14700K as a "125W" processor and the 4070Ti as a "300W" graphics card are very much average numbers and do not represent peaks power draw.

Realistically, there will be moments where the CPU+GPU alone are pulling 280W+360W, some of the transient spikes are likely to be even higher, and that's just those two components - the rest of the hardware in the PC uses power, too!

 

[MightyNerdy93]

its not a prebuild , its a custom build , the psu is msi 750w gen5.
thank you guys all , what about updating the bios version , any lead that could help or its not connected at all to his specific problem ?

 

[Chrispy_]

Without knowing the exact model it's hard to say but MSI make some of the cheapest 750W PSUs on the market, 80+Bronze models that technically meet the spec but aren't necessarily any good. They also make some great PSUs but the point stands that MSI have more cheap offerings than high-quality offerings.

Monitoring software is still the answer. If the PSU turns out to be the weak link, there are two options:

1. buy a better PSU
2. reduce the PL1 and PL2 limits of the processor in the BIOS to, say, 200W - and use MSI afterburner to reduce the GPU power limit by 10% (which will likely only reduce performance by 4-5%)

 

[MightyNerdy93]

Without knowing the exact model it's hard to say but MSI make some of the cheapest 750W PSUs on the market, 80+Bronze models that technically meet the spec but aren't necessarily any good. They also make some great PSUs but the point stands that MSI have more cheap offerings than high-quality offerings.

Monitoring software is still the answer. If the PSU turns out to be the weak link, there are two options:

1. buy a better PSU
2. reduce the PL1 and PL2 limits of the processor in the BIOS to, say, 200W - and use MSI afterburner to reduce the GPU power limit by 10% (which will likely only reduce performance by 4-5%)

MAG A750GL PCIE5​

 

[Chrispy_]

Ok, that's actually a decent PSU.

Cybenetics Gold is essentially a "this PSU doesn't suck" badge. @crmaris is the guy who runs Cybenetics labs and he used to be Techpowerup's PSU reviewer before quitting to focus on his own PSU-testing and certification business.

Get some CPU and GPU temperatures from hardware monitoring logs at the time it next crashes. If it's not a cooling/overheating issue it could also easily be faulty or unstable RAM - so use OCCT's Linpack test for memory, or create a Memtest86 boot drive and run that for an hour to rule out memory errors.

 

[LabRat 891]

Hard Power-Offs under load are typically either:
-Overrunning thermal limits faster than a component can throttle
or
-Power stability/capacity related

 

[Princess Garnet]

Is this a sudden shutdown, or a restart (or seemingly frozen "Black screen")?

In my mind, if it's a shutdown, it suggests a thermal issue or an electrical issue (likely PSU, but it may be a power instability issue with another component).

A restart is more tricky to narrow down as it probably has far more possible causes to check.

Either way, you could look into the WHEA directory and see if there's logs being made there. Same with event viewer. if there's absolutely no clues in either (besides Event ID 41/6008 which don't are merely logs that an unexpected shutdown happened but not cause related), then it further suggests the issue isn't software/driver side.

Also, given that is a Raptor Lake CPU, I would check to see if that may be unstable. If the BIOS isn't the latest, it absolutely should be updated because of that CPU (but I'd probably suggest it as a troubleshooting step regardless). If you're having certain other game/application crashes or error messages, it would suggest the CPU. Namely, crashes during shader compilation (which isn't always relayed to the player is happening) or error messages about being out of memory trying to allocate a render resource (I can't remember the exact message) are infamous tell tale signs of CPU instability/degradation.

I would also test with XMP disabled. If four sticks of memory are present, reduce them to two.

 

[eidairaman1]

Hey guys , didn't where exactly to open this thread so if its not in the right place , i would like to move it please.

so my friend has this rig:
14700k
z790 max
4070ti

recently when my friend plays with me fortnite , the computer just shut down (itsn't BSOD)
so he called the store where he got this pc from and they told him that maybe he needs to change the settings on the menu of power options on the OS and maybe there is a need to replace the power button on the case (4000d)
after 2 days , now it happens again.
what could be the problem ?

What ram, what power supply, what bsod code?