Kernel Power Critical Error in Event Viewer

[alanp1995]

Whenever I launch a game, my system will either power itself down completely or just stay running without a picture.

Upon viewing Windows Events, I am presented with an "Event 41 - Kernel-Power" critical error.

 

Error Details:

"The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly."

 

Could this be more likely a PSU issue or a GPU issue? I tested the PC with a 560Ti without any failure, was able to run benchmarks for 1+ hours. But when it comes to the 980Ti, this issue is consistent.

 

CPU: AMD Ryzen 5 2600 3.4 GHz 6-Core Processor
Motherboard: Gigabyte GA-A320M-H Micro ATX AM4 Motherboard
Memory: Crucial Ballistix 16 GB (2 x 8 GB) DDR4-3200 CL16 Memory
Storage: Kingston A400 240 GB 2.5" Solid State Drive
Storage: Western Digital Caviar Blue 500 GB 3.5" 7200RPM Internal Hard Drive
Video Card: EVGA GeForce GTX 980 Ti 6 GB Superclocked+ ACX 2.0+ Video Card
Case: CiT F3 MicroATX Mini Tower Case
Power Supply: Silverstone Strider Plus 700 W 80+ Bronze Certified Fully Modular ATX

[Tristerin]

Had this issue with a R9 Fury Tri-X - I eliminated GPU sag and it went away.  I hate Event 41, no info at all.  What I read on reddit was that the power delivery pins (due to sag) lose contact and a power failure occurs.  Once I bought a stand and propped it up, Ive never had the issue since.  My board didn't have one of those reinforced slots.  

 

Id DIY a stand and prop that bad boy up, Ill see if I have a pic avail of where I put mine to eliminate sag

[Tristerin]

Old pic (water cooled GPU now) but this was the fix

[.Apex.]

28 minutes ago, alanp1995 said:

Could this be more likely a PSU issue or a GPU issue? I tested the PC with a 560Ti without any failure, was able to run benchmarks for 1+ hours. But when it comes to the 980Ti, this issue is consistent.

Could be related to the PSU, how old is it?

 

I've had a similar issue and it was clearly caused by the PSU when looking at the voltage readings, it was dropping to 11.5-11.6V under load which was reported by my GPU, I'm not sure if the GTX 980Ti has those readings but you can look at the 12V reading of your motherboard instead since they're all on the same rail, software voltage reading isn't that accurate but it can give you an idea of how your PSU is performing, after I upgraded my PSU all the voltage readings are at 12V constantly.

 

GPU sag can also be the culprit as @Tristerin said

 

Use HWiNFO if you want to monitor your system (in Sensors-Only mode)

[alanp1995]

52 minutes ago, Tristerin said:

Had this issue with a R9 Fury Tri-X - I eliminated GPU sag and it went away.  I hate Event 41, no info at all.  What I read on reddit was that the power delivery pins (due to sag) lose contact and a power failure occurs.  Once I bought a stand and propped it up, Ive never had the issue since.  My board didn't have one of those reinforced slots.  

 

Id DIY a stand and prop that bad boy up, Ill see if I have a pic avail of where I put mine to eliminate sag

Wow, if that's the issue then it would be awesome. Cheapest solution of all. My brother actually got the 980Ti at the same time as me and got one of these supports right away. I thought that securing it tightly at the bracket with 2 screws would be enough, but I guess that's just one side. I will get one of these right away. Worth a shot. Thanks. I will report back.

[alanp1995]

28 minutes ago, Syn. said:

Could be related to the PSU, how old is it?

 

I've had a similar issue and it was clearly caused by the PSU when looking at the voltage readings, it was dropping to 11.5-11.6V under load which was reported by my GPU, I'm not sure if the GTX 980Ti has those readings but you can look at the 12V reading of your motherboard instead since they're all on the same rail, software voltage reading isn't that accurate but it can give you an idea of how your PSU is performing, after I upgraded my PSU all the voltage readings are at 12V constantly.

 

GPU sag can also be the culprit as @Tristerin said

 

Use HWiNFO if you want to monitor your system (in Sensors-Only mode)

All solutions I came across were also pointing to the PSU. I nearly bought one earlier, but since the PSU works with the 560Ti and can play games, then why would it be that? That's just my thought process. I'll give a look at voltage readings in the BIOS if possible. Software monitoring for voltages is out of the question since the computer just crashes the second it hits any load.

[Tristerin]

1 hour ago, alanp1995 said:

All solutions I came across were also pointing to the PSU. I nearly bought one earlier, but since the PSU works with the 560Ti and can play games, then why would it be that? That's just my thought process. I'll give a look at voltage readings in the BIOS if possible. Software monitoring for voltages is out of the question since the computer just crashes the second it hits any load.

All solutions I found were pointing at the PSU as well (Corsair HX850) and after discussing (and me arguing with but I was wrong) with @jonnyGURU the issue and nearly replacing my PSU I found that solution in the bowels of REDDIT. 

 

Go outside, get a stick, measure it, break it (lift it a hair higher than it needs to be) and you can see pretty quickly if you have a solution

 

Hell, got lego's?

[.Apex.]

59 minutes ago, alanp1995 said:

All solutions I came across were also pointing to the PSU. I nearly bought one earlier, but since the PSU works with the 560Ti and can play games, then why would it be that? That's just my thought process. I'll give a look at voltage readings in the BIOS if possible. Software monitoring for voltages is out of the question since the computer just crashes the second it hits any load.

I'm not sure if the BIOS reports voltages but that would be pointless anyway since there is no load, I suggest you put the load on the CPU if it's stable that way instead and check the voltages in HWiNFO, you could also limit your CPU usage in Cinebench for example by selecting how many threads you want it to utilize so it won't shut down, but there has to be a load of some kind because the voltage drops happen at load.

 

the 560Ti is a 170W TDP card and the GTX 980 Ti is 250W, that's quite a difference if your PSU is on the edge of stability, my PSU for example didn't have those issues at start and it would keep the voltages consistent at 12V but as it ages the capacitors start to degrade, but again those are only theories the issue could be something else completely that's why I recommend to check the voltages first before buying a PSU and realize that wasn't the issue.

[alanp1995]

1 hour ago, Tristerin said:

All solutions I found were pointing at the PSU as well (Corsair HX850) and after discussing (and me arguing with but I was wrong) with @jonnyGURU the issue and nearly replacing my PSU I found that solution in the bowels of REDDIT. 

 

Go outside, get a stick, measure it, break it (lift it a hair higher than it needs to be) and you can see pretty quickly if you have a solution

 

Hell, got lego's?

Okay so I tested it while supporting it and also lay the PC case sideways to prevent it too and had no luck. Doesn't seem to be the issue. I did however get a benchmark to run for a few seconds. Usually it shuts off right away, so I thought I had a victory for a minute.

 

After the benchmark ran for a few seconds, the card did feel very very warm. Warmer than what I think it should have been after ~5 seconds of benchmarking in Unigine Valley.

 

1 hour ago, Syn. said:

I'm not sure if the BIOS reports voltages but that would be pointless anyway since there is no load, I suggest you put the load on the CPU if it's stable that way instead and check the voltages in HWiNFO, you could also limit your CPU usage in Cinebench for example by selecting how many threads you want it to utilize so it won't shut down, but there has to be a load of some kind because the voltage drops happen at load.

 

the 560Ti is a 170W TDP card and the GTX 980 Ti is 250W, that's quite a difference if your PSU is on the edge of stability, my PSU for example didn't have those issues at start and it would keep the voltages consistent at 12V but as it ages the capacitors start to degrade, but again those are only theories the issue could be something else completely that's why I recommend to check the voltages first before buying a PSU and realize that wasn't the issue.

I guess the next step is to check if the single v12 rail is doing it's job? I'm not sure how putting load on the CPU would verify anything to do with the GPU? The other components can't be the issue since with the 560 Ti, the system runs perfectly. I'm honestly just thinking of replacing the GPU with something like an RX580 8GB, a 185W TDP...

 

 

[alanp1995]

2 hours ago, Syn. said:

I'm not sure if the BIOS reports voltages but that would be pointless anyway since there is no load, I suggest you put the load on the CPU if it's stable that way instead and check the voltages in HWiNFO, you could also limit your CPU usage in Cinebench for example by selecting how many threads you want it to utilize so it won't shut down, but there has to be a load of some kind because the voltage drops happen at load.

 

the 560Ti is a 170W TDP card and the GTX 980 Ti is 250W, that's quite a difference if your PSU is on the edge of stability, my PSU for example didn't have those issues at start and it would keep the voltages consistent at 12V but as it ages the capacitors start to degrade, but again those are only theories the issue could be something else completely that's why I recommend to check the voltages first before buying a PSU and realize that wasn't the issue.

Could I simulate the process using some software on the 560Ti? I am looking at HWInfo but don't see anything related to the 12V rail?...

[.Apex.]

57 minutes ago, alanp1995 said:

I guess the next step is to check if the single v12 rail is doing it's job? I'm not sure how putting load on the CPU would verify anything to do with the GPU? The other components can't be the issue since with the 560 Ti, the system runs perfectly. I'm honestly just thinking of replacing the GPU...

You're only checking if the PSU is keeping the 12V rail stable, and since all of the components are running off of the same 12V rail then it doesn't matter where you put the load on, the higher the wattage the higher the voltage drop will be, so that's why starting with a CPU load should show you at least some result since it crashes with the GPU load, but instead of calling it a GPU load and CPU load let's just say ~65W load (CPU) and 250W load (GPU)

57 minutes ago, alanp1995 said:

After the benchmark ran for a few seconds, the card did feel very very warm. Warmer than what I think it should have been after ~5 seconds of benchmarking in Unigine Valley.

Might wanna check temperatures at idle as well since that would cause instability and shutdowns if it's overheating

[.Apex.]

2 minutes ago, alanp1995 said:

Could I simulate the process using some software on the 560Ti? I am looking at HWInfo but don't see anything related to the 12V rail?...

Did you run HWiNFO in Sensors-only mode? it's a checkbox when you start the program

[alanp1995]

8 minutes ago, Syn. said:

You're only checking if the PSU is keeping the 12V rail stable, and since all of the components are running off of the same 12V rail then it doesn't matter where you put the load on, the higher the wattage the higher the voltage drop will be, so that's why starting with a CPU load should show you at least some result since it crashes with the GPU load, but instead of calling it a GPU load and CPU load let's just say ~100W load (CPU) and 250W load (GPU)

Might wanna check temperatures at idle as well since that would cause instability and shutdowns if it's overheating

So I managed to get the voltage reading, it goes between 11.448V and 11.880V on the 12V rail. Seems perfect to me.

The GPU idles at 38C, I reapplied the thermal paste earlier today and knocked 6C off it at idle.

Again I can't get a reading for GPU temps under load as I can't push any load with it.

[.Apex.]

1 minute ago, alanp1995 said:

So I managed to get the voltage reading, it's a steady 11.880V on the 12V rail. Seems perfect to me.

The GPU idles at 38C, I reapplied the thermal paste earlier today and knocked 6C off it at idle.

Again I can't get a reading for GPU temps under load as I can't push any load with it.

is that 11.88V reading at idle or are you loading up the CPU? either way that's pretty bad..

[alanp1995]

1 minute ago, Syn. said:

is that 11.88V reading at idle or are you loading up the CPU? either way that's pretty bad..

It's actually between 11.448V and 11.880V. That reading was in the BIOS. I managed to use HWInfo to create a detailed log including the 12V readings. Tomorrow I will put back in the 980 Ti and start the logging before running a benchmark, it should output what's needed even if the PC crashes right?

[.Apex.]

7 minutes ago, alanp1995 said:

It's actually between 11.448V and 11.880V. That reading was in the BIOS. I managed to use HWInfo to create a detailed log including the 12V readings. Tomorrow I will put back in the 980 Ti and start the logging before running a benchmark, it should output what's needed even if the PC crashes right?

around 11.4-11.5 is the absolute limit of the ATX specification, but no decent PSU would let the voltages drop that low because it will cause severe instability issues, It's pretty safe to say that your PSU is the issue

 

It should output the result before it crashes I think, but you still haven't mentioned if you put any load on the CPU

[alanp1995]

2 minutes ago, Syn. said:

around 11.4-11.5 is the absolute limit of the ATX specification, but no decent PSU would let the voltages drop that low because it will cause severe instability issues, It's pretty safe to say that your PSU is the issue

 

It should output the result before it crashes I think, but you still haven't mentioned if you put any load on the CPU

I will put load on the CPU and let run for a bit, then begin a GPU benchmark and see what the readings are like.

Am I still looking at the 12V for the CPU load? Not sure what I should be looking for when loading the CPU...

[.Apex.]

5 minutes ago, alanp1995 said:

I will put load on the CPU and let run for a bit, then begin a GPU benchmark and see what the readings are like.

Am I still looking at the 12V for the CPU load? Not sure what I should be looking for when loading the CPU...

It's not necessary to let it run for a bit since the voltage drop will happen instantly, and yeah you're still looking at the 12V reading in HWiNFO, your CPU and GPU and Motherboard all run from the same 12V rail, if a voltage drop happens it will happen across all of the components

 

That specific 12V reading you're looking at is taken from the Motherboard but as I mentioned that will translate to the CPU and GPU as well.

[.Apex.]

3 minutes ago, alanp1995 said:

I will put load on the CPU and let run for a bit, then begin a GPU benchmark and see what the readings are like.

Am I still looking at the 12V for the CPU load? Not sure what I should be looking for when loading the CPU...

Cinebench is easy to use if you wanna stress the CPU

[alanp1995]

15 minutes ago, Syn. said:

It's not necessary to let it run for a bit since the voltage drop will happen instantly, and yeah you're still looking at the 12V reading in HWiNFO, your CPU and GPU and Motherboard all run from the same 12V rail, if a voltage drop happens it will happen across all of the components

 

That specific 12V reading you're looking at is taken from the Motherboard but as I mentioned that will translate to the CPU and GPU as well.

Just ran Prime95 for 15 minutes. Ran into no problems and CPU temps were great. The 12V results averaged around 11.808V.

 

Here is a link to the CSV Log for HWInfo - https://drive.google.com/file/d/1_bY056OTKKRZNlj7WzQ_ybPOzH3lgBK9/view?usp=sharing

 

 

[.Apex.]

22 minutes ago, alanp1995 said:

Just ran Prime95 for 15 minutes. Ran into no problems and CPU temps were great. The 12V results averaged around 11.808V.

 

Here is a link to the CSV Log for HWInfo - https://drive.google.com/file/d/1_bY056OTKKRZNlj7WzQ_ybPOzH3lgBK9/view?usp=sharing

 

 

Yeah this is pretty bad, if with a 65W load from the CPU and it's already dropping to 11.59V then there's no way it can handle the GPU at 250W, it can be unsafe to keep the PC running, if the PSU blows out it might take your whole PC with it, since the PSU was not intended to run at this voltage by design then it's most likely faulty/breaking down, I suggest you keep your PC off until you get a new PSU

[.Apex.]

12 minutes ago, Syn. said:

Yeah this is pretty bad, if with a 65W load from the CPU and it's already dropping to 11.59V then there's no way it can handle the GPU at 250W, it can be unsafe to keep the PC running, if the PSU blows out it might take your whole PC with it, the PSU was not intended to run at this voltage so it's most likely faulty/breaking down, I suggest you keep your PC off until you get a new PSU

Also don't cheap out on the PSU if you want it to last longer and want it to be able to keep its voltage rails stable

[.Apex.]

10 minutes ago, Syn. said:

if the PSU blows out it might take your whole PC with it

By blows out I mean by a power surge caused by shorting components, not a literal explosion, though the capacitors can blow up but it's unlikely I think.. there are many cases of fires caused by PSUs though that's why it's important to get a quality PSU

[alanp1995]

9 hours ago, Syn. said:

By blows out I mean by a power surge caused by shorting components, not a literal explosion, though the capacitors can blow up but it's unlikely I think.. there are many cases of fires caused by PSUs though that's why it's important to get a quality PSU

Yeah I understand, a power surge would be the last thing I want. I was thinking of getting the Corsair TX550M 80+ Gold PSU, it seems to be the way to go when it comes to sub €100 choices. It's weird how within a matter of hours, the PSU just degraded like that after using it perfectly fine for over 2 months with the 980Ti.

[.Apex.]

1 hour ago, alanp1995 said:

Yeah I understand, a power surge would be the last thing I want. I was thinking of getting the Corsair TX550M 80+ Gold PSU, it seems to be the way to go when it comes to sub €100 choices. It's weird how within a matter of hours, the PSU just degraded like that after using it perfectly fine for over 2 months with the 980Ti.

The TX550M would be a great choice, and the PSU might have had a defect from factory if it's only 2 months old I'm not sure

Here's a review from Amazon for the PSU, it seems like you're not the only one that dealt with this:

Quote

"OK, spent some money buying a "decent" modular power supply (Silverstone SST-ST75F-GS Strider Gold Series modular). First one failed miserably after a few months, replacement took way too long for SCAN UK to send out. Finally got another, and now replacement failed. Unfortunately, this POS is now out of its short warranty and I had noted that the power was never stable, causing reboots ( I know this as I moved the P/S to another machine and it did the same), and is a severe disappointment from Silverstone. So, another useless brand to cross off my buying list, if two fail consecutively, it's enough for me to NEVER BUT SILVERSTONE AGAIN!
So, if you're thinking of buying one of these, I'd suggest you avoid it. Also, if you're thinking of purchasing from SCAN UK, also avoid them too, I've had too many issues with them and their poor customer service and obviously warranty replacement parts."