pc restarts while playing games or doing a stress test

[lavlav]

My computer restarts when i playing games or doing a stress test to the psu.
I sent the power supply for testing in the store and it turned out that there is no problem.

i have:
sapphire rx 6900 xt
ryzen 9 5900x
32GB ddr4 3600
asus rog strix 850w 80+ gold

also i have a new cpu, Motherboard and ram

do you know what the problem is?

 

[xtreemchaos]

could be your ram mate, id drop it down to 3200 just to see and if it dont crash then go to 3400. do you have the latest bios for your motherboard ?.

 

[64K]

My computer restarts when i playing games or doing a stress test to the psu.
I sent the power supply for testing in the store and it turned out that there is no problem.

i have:
sapphire rx 6900 xt
ryzen 9 5900x
32GB ddr4 3600
asus rog strix 850w 80+ gold

also i have a new cpu, Motherboard and ram

do you know what the problem is?

Despite what they told you at the store I still suspect your PSU. Do you have access to another PSU that you could swap in and see if you continue to have this problem with a different PSU?

 

[ThrashZone]

My computer restarts when i playing games or doing a stress test to the psu.
I sent the power supply for testing in the store and it turned out that there is no problem.

i have:
sapphire rx 6900 xt
ryzen 9 5900x
32GB ddr4 3600
asus rog strix 850w 80+ gold

also i have a new cpu, Motherboard and ram

do you know what the problem is?

Hi,
Get a kill-a-watt version depending on region

https://www.amazon.com/P3-P4400-Electricity-Usage-Monitor/dp/B00009MDBU

Test wall outlets to.

 

[Bill_Bright]

Check your temps and make sure the interior is clean of heat trapping dust.

 

[ir_cow]

The question is, did the store really do a full load test, or just turn it on?

 

[Lei]

Check your temps and make sure the interior is clean of heat trapping dust.

the temps are very low. under 60/70.
i have a custom water cooling system

 

[Hofnaerrchen]

could be your ram mate, id drop it down to 3200 just to see and if it dont crash then go to 3400. do you have the latest bios for your motherboard ?.

RAM being the problem you'll rather get a BSOD - something the threadstarter has not reported so far. Restarting without further entries - except the system shutting down suddenly - to the event log is probably caused because of power issues - not neccessarily the PSU. First thing to do: Change BIOS settings for sudden power loss - this way you'll get a first impression what's wrong. Maybe just a power cable is loose or missing - like the seperate 8 pin connector. Another question: Do you use different PCIe-power cables or a split one. Don't know the ASUS PSU but maybe you'll just have to adjust the power load to different rails - check the manual for that. I would start to check all connections and then you can start stress testing components one by one.

 

[Lei]

Do you use different PCIe-power cables or a split one. Don't know the ASUS PSU but maybe you'll just have to adjust the power load to different rails - check the manual for that. I would start to check all connections and then you can start stress testing components one by one.

Yeah, that's a very good idea. Plumb in a separate cable for each pin sockets. @lavlav

 

[dirtyferret]

The question is, did the store really do a full load test, or just turn it on?

I believe that is what you call a rhetorical affirmation

 

[r9]

My computer restarts when i playing games or doing a stress test to the psu.
I sent the power supply for testing in the store and it turned out that there is no problem.

i have:
sapphire rx 6900 xt
ryzen 9 5900x
32GB ddr4 3600
asus rog strix 850w 80+ gold

also i have a new cpu, Motherboard and ram

do you know what the problem is?

You need to rule out one thing at a time.
If it restarts while CPU stress test you can rule out the GPU.
Make sure the RAM timings are correct and run mem test for the RAM or do CPU stress test with one RAM stick at a time.
If can't find pattern doing the above could be the PSU or motherboard.

 

[64K]

RAM being the problem you'll rather get a BSOD - something the threadstarter has not reported so far. Restarting without further entries - except the system shutting down suddenly - to the event log is probably caused because of power issues - not neccessarily the PSU. First thing to do: Change BIOS settings for sudden power loss - this way you'll get a first impression what's wrong. Maybe just a power cable is loose or missing - like the seperate 8 pin connector. Another question: Do you use different PCIe-power cables or a split one. Don't know the ASUS PSU but maybe you'll just have to adjust the power load to different rails - check the manual for that. I would start to check all connections and then you can start stress testing components one by one.

Wouldn't it be a good idea to first eliminate the possibility that it is simply a PSU going bad before trying a lot of complicated fixes? That is why I suggested swapping the PSU out for another to see if the problem goes away.

 

[Hofnaerrchen]

Wouldn't it be a good idea to first eliminate the possibility that it is simply a PSU going bad before trying a lot of complicated fixes? That is why I suggested swapping the PSU out for another to see if the problem goes away.

For this reason I suggested to change BIOS power loss settings and checking the event log. It's most useful to replicate the problem by changing software settings first and checking cables and maybe even how well other components fit. As he did mention to have made some changes to the system it could also be such a simple thing than just a lose screw. So checking the build for this is also more useful than "simply" replacing a component just because someone is making a not very educated (based on limited information) guess of the problem on a forum.

 

[64K]

For this reason I suggested to change BIOS power loss settings and checking the event log. It's most useful to replicate the problem by changing software settings first and checking cables and maybe even how well other components fit. As he did mention to have made some changes to the system it could also be such a simple thing than just a lose screw. So checking the build for this is also more useful than "simply" replacing a component just because someone is making a not very educated (based on limited information) guess of the problem on a forum.

Never suggested replacing PSU. Just swapping it with another to eliminate the possibility that it is the PSU. Usually the symptoms described are a PSU going bad. If that gets eliminated as a possibility then it's time to proceed with further fixes.

 

[lavlav]

Thank you. I will try everything and come back with an answer.

When I run a stress test for the RAM or the CPU, nothing happens.

This happens mainly when I run a stress test for the power supply (all parts 100%)

The store did a full inspection for several hours and found no problem.

I also have cutom extension cables that I use and I didn't give them to the store, could that be the problem?

The power supply is connected in:
2*8 pin to GPU
2*8 pin to the motherboard

Never suggested replacing PSU. Just swapping it with another to eliminate the possibility that it is the PSU. Usually the symptoms described are a PSU going bad. If that gets eliminated as a possibility then it's time to proceed with further fixes.

I am looking for another power supply for testing

 

[Hofnaerrchen]

Thank you. I will try everything and come back with an answer.

When I run a stress test for the RAM or the CPU, nothing happens.

This happens mainly when I run a stress test for the power supply (all parts 100%)

The store did a full inspection for several hours and found no problem.

I also have cutom extension cables that I use and I didn't give them to the store, could that be the problem?

The power supply is connected in:
2*8 pin to GPU
2*8 pin to the motherboard


I am looking for another power supply for testing

Did the store use different cables or those you use? Did the store use a completely different system or yours to do the inspection?

As I already said: Could be just using a wrong cable setup, maybe your custom cables - try replacing them with the ones that came with the PSU before replacing the PSU.

I use a system that is comparable to yours: 5800X with a XFX 6900 XT (also two 8-pin cables) on a be quiet 850W PSU. In my case - check your PSU's manual - it was stated to plug PCIe power cables in a specific manner to prevent overloading the rails - which could be the reason your system restarts - which it only does when there is load on CPU and GPU as you said.

Edit: If I remember correctly techpowerup got power spikes of 600+W when testing the 6900 XT. If this is the case with your GPU (silicon lottery) and it's not the cabling, it does not necessarily mean your PSU is damaged/broken, it might just do what it is supposed to do if this happens: Trigger overcurrent protection. but in this case you either have to reduce the PPT of your GPU and/or CPU or actually get a better (1000W) and/or higher quality PSU, maybe a PSU of a different brand.

Edit 2: Just out of curiosity I did check the ASUS page, there is a "RECOMMENDED PSU TABLE". Based on that table ASUS recommends using their 1000W PSU when combining an R9 with the 6900 XT. Might answer the question why your system restarts when under full GPU and CPU load.

 

[lavlav]

Did the store use different cables or those you use? Did the store use a completely different system or yours to do the inspection?

As I already said: Could be just using a wrong cable setup, maybe your custom cables - try replacing them with the ones that came with the PSU before replacing the PSU.

I use a system that is comparable to yours: 5800X with a XFX 6900 XT (also two 8-pin cables) on a be quiet 850W PSU. In my case - check your PSU's manual - it was stated to plug PCIe power cables in a specific manner to prevent overloading the rails - which could be the reason your system restarts - which it only does when there is load on CPU and GPU as you said.

Edit: If I remember correctly techpowerup got power spikes of 600+W when testing the 6900 XT. If this is the case with your GPU (silicon lottery) and it's not the cabling, it does not necessarily mean your PSU is damaged/broken, it might just do what it is supposed to do if this happens: Trigger overcurrent protection. but in this case you either have to reduce the PPT of your GPU and/or CPU or actually get a better (1000W) and/or higher quality PSU, maybe a PSU of a different brand.

Edit 2: Just out of curiosity I did check the ASUS page, there is a "RECOMMENDED PSU TABLE". Based on that table ASUS recommends using their 1000W PSU when combining an R9 with the 6900 XT. Might answer the question why your system restarts when under full GPU and CPU load.

The store use the same cable but without the custom extension cable.
they used a different system with 3080

I read a lot on the internet and they say that 850 is completely enough for what I have.

 

[maxfly]

Eliminate the extension and re-run your games or tests. Most likely your either running into a poor extension cable or transient spikes are tripping your PSUs overcurrent protection. Probably the latter.
The shop should have tested your card, in your rig as a whole (with the extension present). Their testing only verified that your PSU works with their gear.

 

[AlwaysHope]

The store use the same cable but without the custom extension cable.
they used a different system with 3080

I read a lot on the internet and they say that 850 is completely enough for what I have.

OP, you are correct, a quality 850w psu is adequate for your hardware. The question is like what @Hofnaerrchen was referring to in that post, is to be absolutely sure the correct PCIe power cables are in the correct device end sockets.
As a reference point, a 700w PSU is recommended for an RX 6800 XT system like my gaming rig but considering your RX 6950 XT has the same power requirements as evidenced in the TPU database for these cards, then I can only suggest being sure that the right power cables are in the right sockets - I can't stress this enough, just check & be double sure. I've never seen my system when fully OC'ed in benching for example or intense gaming ever exceed a total maximum peak of 550w approx as measured from the wall socket with an electricity measurement device & that's using a renowned power hungry rocket lake i7 platform which of course will drink more power than your 5900X. So I do not think your 850w PSU is inadequate.

 

[TheDeeGee]

Could be a DOA 5900X, last year i also had the biggest struggle with my first ever AMD system in years, and it turned out the 5800X i had was simply defect and NOTHING would make it stable.

My 11700 been running crash free for 19 months now.

 

[lavlav]

I have a custom water cool system.
it means the my gpu as a water block so the warranty is no longer available.

I will check the extension cables.

Also i only gave the store the power supply without the extension cables.

 

[OkieDan]

I use a system that is comparable to yours: 5800X with a XFX 6900 XT (also two 8-pin cables) on a be quiet 850W PSU. In my case - check your PSU's manual - it was stated to plug PCIe power cables in a specific manner to prevent overloading the rails - which could be the reason your system restarts - which it only does when there is load on CPU and GPU as you said.

Couldn't find full manual for this only quick start guide which didn't mention anything about overloading rails. It's a single rail PSU unlike yours, so it would make sense not needing to worry about this on OP's PSU.

The store use the same cable but without the custom extension cable.

I typically try to rule out as much as possible no matter how unlikely, starting with the easiest things first unless a more likely possibility is glaringly obvious.

 

[Hofnaerrchen]

Could be a DOA 5900X, last year i also had the biggest struggle with my first ever AMD system in years, and it turned out the 5800X i had was simply defect and NOTHING would make it stable.

My 11700 been running crash free for 19 months now.

Very unlikely. lavlav said he has no problems when stressing the CPU. It just happens when the whole system is under full load.

 

[lavlav]

 

[OkieDan]

View attachment 267409

Looks fine, can always try swapping gpu with cpu as a troubleshooting step even though it shouldn't matter.