[SOLVED] Kernal Power Even ID 41 (random reboots)

[kimbo_0331]

Hello,

Recently I just assembled a new gaming PC (See specs below). After about 3 weeks of gameplay, when I transitioned from a 2080ti to a 4090, I started having random reboots during gameplay only. Right now it has only occurred while playing Warhammer 3, and It Takes Two. At first I though it was software conflict because it was limited to one title, but after it started happening to a second game, I started to get worried all over again. I have ran Furmark (20 min) with no issues. I have played Warzone 2.0 with no issues (1 hr). I have played HL Alyx with no issues (2 hrs). I played Portal RTX with no issues (1 hr). It seems to be limited to both warhammer 3 (about 10 min) and It Takes Two for now (about 15 min). I will be purchasing A Plague Tale Requiem which from my research has also been a known source to duplicate this observation.

I have opened my event viewer to reviewer the Administrative Events at the time of the reboot. I get the Kernal0Power Event ID 41 ("the system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly). In the system information of the windows Logs at the time of the reboot it has Kernel Power Even ID 125 (ACPI thermal zone \_TZ.TZ00 has been enumerated).

I don't know what the problem could be. Right now I am leaning towards a faulty GPU or a faulty PSU. From the research I have conducted, others are having issues with the Thermaltake GF3 1650w PSU, but I don't want to complete an RMA without knowing for sure. To do this I ran the CPUID hardware monitor, but I am not tech savvy enough to discern any anomalies (see below for pictures). I can tell based on temps that it cannot be an overheating issue with the GPU, or the CPU which were my initial concerns. See below for those results after a recent unexpected reboot while playing warhammer 3.

If anyone is able to offer any insight onto a fix for this obersavtion, or at a minimum, provide some strong leads, I will issue a reward in form of a smile. After I have fixed the issue I will engage in celebratory gunfire and other festivities.

Things i've done to help troubleshoot based on what others have suggested:

• updated GPU drivers
• reintalled GPU drivers via Geforce Experience
• Used a DDU to erase all data pertaining to the 4090 and reinstalled drivers
• updated windows
• Turned off automatic restart
• Changed power options to balanced
• Changed power options to performance
• used a 2080ti (no random reboots)
• ran windows memory diagnostic
• disabled XMP
• enabled XMP
• ran furmark

System (Specs):
-Intel Core i7-13700K 3.4 GHz 16-Core Processor
-NZXT Kraken Z73 RGB 52.44 CFM Liquid CPU Cooler
-MSI MPG Z790 EDGE WIFI ATX LGA1700 Motherboard
-OLOy Blade RGB 32 GB (4 x 8 GB) DDR5-5600 CL40 Memory
-Samsung 980 Pro 2 TB M.2-2280 PCIe 4.0 X4 NVME Solid State Drive
-Samsung 980 Pro 2 TB M.2-2280 PCIe 4.0 X4 NVME Solid State Drive
-MSI SUPRIM X GeForce RTX 4090 24 GB Video Card
-Thermaltake Toughpower GF3 TT Premium 1650 W 80+ Gold Certified Fully Modular ATX Power Supply
-Microsoft Windows 11 Pro OEM - DVD 64-bit
-Lian Li Uni Fan SL-Infinity 61.3 CFM 120 mm Fan
-Lian Li Uni Fan SL-Infinity 61.3 CFM 120 mm Fans 3-Pack
-Lian Li Uni Fan SL-Infinity 61.3 CFM 120 mm Fans 3-Pack
-Lian Li Uni Fan SL-Infinity 61.3 CFM 120 mm Fans 3-Pack
-Lian Li O11 Dynamic EVO ATX Mid Tower Case

Thanks,

 

[Aeacus]

While your PSU has enough wattage, it should keep RTX 4090 running. But fact is, it doesn't. So, test with 2nd, good quality and known to work PSU, to see if your issues go away. Most likely they will.

For 2nd PSU, i suggest getting one of the 1600W "kings",
pcpp: https://pcpartpicker.com/products/compare/cJbwrH,2BcRsY,tFvdnQ/

As of why i think issue is with PSU:

1. RTX 2080 Ti (250W) runs fine, while RTX 4090 (450W) does not.
2. Thermaltake isn't my 1st choice when it comes to PSUs. It isn't even 3rd choice. Meaning Tt PSUs, most of them, are sub-par.

 

[vladakv]

Check for motherboard broken/distorted capacitor. Check socket pins... I use to have old motherboard which has broken capacitor and i got a lot of random reboots with kernel event 41 critical error.
I hope its something not serious in your case.

 

[Aeacus]

I use to have old motherboard

OP has Z790 chipset MoBo. Is that "old" for you?

 

[kimbo_0331]

While your PSU has enough wattage, it should keep RTX 4090 running. But fact is, it doesn't. So, test with 2nd, good quality and known to work PSU, to see if your issues go away. Most likely they will.

For 2nd PSU, i suggest getting one of the 1600W "kings",
pcpp: https://pcpartpicker.com/products/compare/cJbwrH,2BcRsY,tFvdnQ/

As of why i think issue is with PSU:

1. RTX 2080 Ti (250W) runs fine, while RTX 4090 (450W) does not.
2. Thermaltake isn't my 1st choice when it comes to PSUs. It isn't even 3rd choice. Meaning Tt PSUs, most of them, are sub-par.

Aeacus,

Thanks for you input. I have to agree with you about thermaltake. I know that they make good products, but there PSU's seem to be lacking. I have seen multiple threads of persons having similar issues with the PSU's.

I will add that I contacted Microsoft support yesterday who manage to find some corrupt drivers with windows 11. Those were fixed coupled with a Windows 11 repair. So far, there have been no crashes. I'm not going to call the issue fixed. I want to conduct some extensive testing first.

Thanks,

 

[kimbo_0331]

Check for motherboard broken/distorted capacitor. Check socket pins... I use to have old motherboard which has broken capacitor and i got a lot of random reboots with kernel event 41 critical error.
I hope its something not serious in your case.

Vladakv,
Thanks for the input. If this were a hardware issue with my motherboard, I would expect to see similar symptoms regardless of what GPU or PSU i have opperating. However, I will keep this in mind. Thanks for helping me out brotha.

 

[vladakv]

OP has Z790 chipset MoBo. Is that "old" for you?

Not this one. I mean old - Asus P5 WS Evolution like 10 years ago 🫤

 

[Aeacus]

Not this one. I mean old - Asus P5 WS Evolution like 10 years ago 🫤

And why exactly you apply the old hardware issue to brand new hardware?

 

[vladakv]

And why exactly you apply the old hardware issue to brand new hardware?

Maybe he did some mistake during build. Could be some damaged component of the mobo. What do you think?

 

[Aeacus]

Maybe he did some mistake during build. Could be some damaged component of the mobo. What do you think?

In a slim chance that you'll knock capacitor loose or completely off the board (yes, it has happened), will result that system won't get past POST. Let alone booting into OS.

Only change in OPs build, was GPU.
Build works fine with RTX 2080 Ti but had issues with RTX 4090.

Symptoms wise, there are 3 reasons why PC suddenly powers off or reboots:

1. Way too high CPU/GPU temps.
2. Inadequate PSU, which fails to keep system running on high load.
3. Electricity grid issues.

#1 OP has already established that neither his CPU or GPU thermal throttles. So, high temps aren't the reason.

#2 While 1650W is more than enough for RTX 4090 build, the Thermaltake PSU isn't known for good build quality. On the contrary, there are several cases where people have had issues with Thermaltake PSU, even the very same one as OP has. So, issue being with PSU is very likely. Especially since using 250W GPU has system running fine, while when using 450W GPU, issues arise.

#3 Main electricity grid can be an issue as well (blackout, brownout, surges). But then it would be random and doesn't matter what load PC has on it. But since OP has established certain instances when system powers off, it is very unlikely that electricity grid would be an issue. Still, getting an UPS will prevent all and any issues that might come from electricity grid and running that expensive of a hardware - UPS is a must.

 

[kimbo_0331]

While your PSU has enough wattage, it should keep RTX 4090 running. But fact is, it doesn't. So, test with 2nd, good quality and known to work PSU, to see if your issues go away. Most likely they will.

For 2nd PSU, i suggest getting one of the 1600W "kings",
pcpp: https://pcpartpicker.com/products/compare/cJbwrH,2BcRsY,tFvdnQ/

As of why i think issue is with PSU:

1. RTX 2080 Ti (250W) runs fine, while RTX 4090 (450W) does not.
2. Thermaltake isn't my 1st choice when it comes to PSUs. It isn't even 3rd choice. Meaning Tt PSUs, most of them, are sub-par.

Aeacus,

You know I kind of thought that thermaltake was a good brand. This is actually the first PSU I have purchased from them. I usually get EVGA. Because I needed an ATX 3.0 PSU with a 16 pin socket, that pretty much limited my options to a few brands.

I will say that after the windows technician helped to reinstall windows 11 that the crashes stopped, or so I thought. Warhammer 3 would reboot my PC after 10 to 20 minutes of gameplay. I playtested it for 45 and it didn't crash, but that was also the last time I checked. It takes two would crash after 30 min of gameplay. I played for two hours before the reboot happended again. This is how I knew the problem had not been fixed

I do have to wonder that why a windows 11 repair would make the crashes go from 20 to 30 min to 2 hours. This makes me wonder if it's not a software issues. What are your thoughts? Do you still think it is the PSU. Thermaltake has already volunteered to do an RMA.

Thanks for your input,

 

[Aeacus]

You know I kind of thought that thermaltake was a good brand.

Thermaltake makes good PC cases. Also, they innovate with PC cases and some of the most unique PC cases also come from them. PSU wise, they aren't that good. Sure, there are few good models (like DPS G), but bulk of their units are sub-par.

With EVGA, you also need to be careful. EVGA also has some good units, like P2 and T2. But they also have a lot of sub-par units, including the latest G5 series.

With PSU, i'd stick to Seasonic or Super Flower. Since Seasonic and Super Flower are two of the best three PSU OEMs out there (the 3rd one being Flextronics, who made Corsair AXi series).
Further reading: https://forums.tomshardware.com/thr...pply-discussion-thread-toms-hardware.3212332/

I do have to wonder that why a windows 11 repair would make the crashes go from 20 to 30 min to 2 hours. This makes me wonder if it's not a software issues. What are your thoughts? Do you still think it is the PSU. Thermaltake has already volunteered to do an RMA.

I don't run Win11 and can't tell the diff between Win10 and Win11. For that, i'd need to summon our master of Win11: @Colif , who should be able to tell why is it like so.

I'm specialized on hardware issues and unless you test with 2nd PSU (preferably another brand), we can not rule out PSU issue. Btw, PSU RMA may not give any good results, if the issue is with the specific platform, made by High Power. (High Power is the OEM of your Thermaltake unit and High Power doesn't have good rep among PSU OEMs.)

Btw, your PSU doesn't meet ATX PSU standard, set in place for all PSU OEMs to follow, so that PSUs are safe to use. Namely the hold-up time is 11.3ms, while at minimum, it must be at least 17ms.
Your PSU review: https://www.techpowerup.com/review/thermaltake-toughpower-gf3-1650-w/8.html

For comparison, i'm using Seasonic PRIME 650 80+ Titanium [SSR-650TD] PSU, and it's hold up time is mythical 35.3ms. That's double of what ATX PSU standard specifies it to be + then some. Then again, i bought the best 650W PSU money could buy.
My PSU review: https://www.tomshardware.com/reviews/seasonic-prime-titanium-650w-psu,4690-4.html

By the looks of it, your PSU is the cheapest 1.6 kW unit you could buy. That much we can read from the review as well. And if you pay peanuts - you will get monkeys.

 

[Karadjgne]

Couple of assumptions.

First being that the 4090 is a 450w card. Doubtful it's hooked up that way on a 1650w psu. More likely it's hooked up to be a 600w, so able to pull closer to 500w power limits. That means spikes closer to 800w range.

Second, that's a TT 1650w. And it's Gold rated. Right off the bat that's a disaster, that kind of wattage has no business being that inefficient. Flat out you'd be looking at spike values for the entire pc of @ 1100-1200w, which is close to 1400w from the wall, or @ 11-11.5A just for the pc, not counting monitors, lights, the alarm clock, iPad chargers etc.

Don't turn on the ceiling fan, you just starved the pc... Psus in the 1600w+ class should be a Minimum of Platinum rated with Titanium preferred. You'd not want a 850w Bronze in a pc would you?

1600w+ psus should be on a dedicated 15A circuit, or at least a shared 20A circuit if total power consumption at any time is anywhere close to psu limits.

That's also assuming the circuit is actually good, I've seen many houses with lackluster wiring connections, stabbed outlets half cooked that couldn't handle the full 15A rated draw without tripping the breaker or burning connections up.

 

[kimbo_0331]

that inefficient

Thermaltake makes good PC cases. Also, they innovate with PC cases and some of the most unique PC cases also come from them. PSU wise, they aren't that good. Sure, there are few good models (like DPS G), but bulk of their units are sub-par.

With EVGA, you also need to be careful. EVGA also has some good units, like P2 and T2. But they also have a lot of sub-par units, including the latest G5 series.

With PSU, i'd stick to Seasonic or Super Flower. Since Seasonic and Super Flower are two of the best three PSU OEMs out there (the 3rd one being Flextronics, who made Corsair AXi series).
Further reading: https://forums.tomshardware.com/thr...pply-discussion-thread-toms-hardware.3212332/



I don't run Win11 and can't tell the diff between Win10 and Win11. For that, i'd need to summon our master of Win11: @Colif , who should be able to tell why is it like so.

I'm specialized on hardware issues and unless you test with 2nd PSU (preferably another brand), we can not rule out PSU issue. Btw, PSU RMA may not give any good results, if the issue is with the specific platform, made by High Power. (High Power is the OEM of your Thermaltake unit and High Power doesn't have good rep among PSU OEMs.)

Btw, your PSU doesn't meet ATX PSU standard, set in place for all PSU OEMs to follow, so that PSUs are safe to use. Namely the hold-up time is 11.3ms, while at minimum, it must be at least 17ms.
Your PSU review: https://www.techpowerup.com/review/thermaltake-toughpower-gf3-1650-w/8.html

For comparison, i'm using Seasonic PRIME 650 80+ Titanium [SSR-650TD] PSU, and it's hold up time is mythical 35.3ms. That's double of what ATX PSU standard specifies it to be + then some. Then again, i bought the best 650W PSU money could buy.
My PSU review: https://www.tomshardware.com/reviews/seasonic-prime-titanium-650w-psu,4690-4.html

By the looks of it, your PSU is the cheapest 1.6 kW unit you could buy. That much we can read from the review as well. And if you pay peanuts - you will get monkeys.

Aeacus,

Honestly, most of the PSU specifications go above my head. I just have to take your word for it. Are you saying that the Thermaltake Toughpower GF3 1650W PSU is substandard for a my PC build?

Also, can you elaborate more on what you mean when you say "our PSU doesn't meet ATX PSU standard, set in place for all PSU OEMs to follow, so that PSUs are safe to use. Namely the hold-up time is 11.3ms, while at minimum, it must be at least 17ms."

Thansk,

 

[kimbo_0331]

Couple of assumptions.

First being that the 4090 is a 450w card. Doubtful it's hooked up that way on a 1650w psu. More likely it's hooked up to be a 600w, so able to pull closer to 500w power limits. That means spikes closer to 800w range.

Second, that's a TT 1650w. And it's Gold rated. Right off the bat that's a disaster, that kind of wattage has no business being that inefficient. Flat out you'd be looking at spike values for the entire pc of @ 1100-1200w, which is close to 1400w from the wall, or @ 11-11.5A just for the pc, not counting monitors, lights, the alarm clock, iPad chargers etc.

Don't turn on the ceiling fan, you just starved the pc... Psus in the 1600w+ class should be a Minimum of Platinum rated with Titanium preferred. You'd not want a 850w Bronze in a pc would you?

1600w+ psus should be on a dedicated 15A circuit, or at least a shared 20A circuit if total power consumption at any time is anywhere close to psu limits.

That's also assuming the circuit is actually good, I've seen many houses with lackluster wiring connections, stabbed outlets half cooked that couldn't handle the full 15A rated draw without tripping the breaker or burning connections up.

Karadjgne,

Thanks for your input. It sounds like based on what you and your assoicates have said is that I bought a fancy paper weight. What PSU would you recommend for a 4090 that is ATX 3 copatible with the 16 pin power connector?

Thanks,

 

[Colif]

I will just add software tests like CPUID aren't accurate for testing a PSU on. Only real accurate test is a multimeter.

Event 41 (63) is reported by windows after an unexpected restart. It doesn't necessarily mean its the PSU. It can be, but its not an obvious sign. Windows used to mistakenly report one if you used fast startup.

@Aeacus lucky I even noticed this, not really here at the moment.

 

[Karadjgne]

Corsair HXi 1500w. Can get the pcie cable directly from Corsair for $20 if it's not included.

Seasonic Vertex 1200w, beQuiet Dark Power Pro 12 and others are atx 3.0 compliant.

With your pc, anything above 1200w isn't really necessary as the pc can't use that much power on its best day and ultimate load. Gaming loads are typically right around 70% of whatever the pc max is, at most. So you'd be @ 800w range, with the occasional spike maybe hitting 1000w for @ 10ms.

 

[Aeacus]

Are you saying that the Thermaltake Toughpower GF3 1650W PSU is substandard for a my PC build?

Not all PSUs are created equal.

For example, can you tell me why your PSU costs easy $200 less than other 1.6kW PSUs,
pcpp: https://pcpartpicker.com/products/compare/LbtWGX,cJbwrH,tFvdnQ,X93gXL/


Sure, Seasonic and Corsair PSUs are 80+ Titanium (highest efficiency) but EVGA PSU is 80+ Platinum, which is only 2% more efficient over 80+ Gold. Yet, even the EVGA PSU costs 200 bucks more than your PSU.

Or another example,
pcpp: https://pcpartpicker.com/products/compare/RzfFf7,n34BD3/

Can you tell, why Seasonic PSU costs 50 bucks more than Enermax unit? Both are 650W, fully-modular and 80+ Gold efficiency. And which of the two would you get?

Also, can you elaborate more on what you mean when you say "our PSU doesn't meet ATX PSU standard, set in place for all PSU OEMs to follow, so that PSUs are safe to use. Namely the hold-up time is 11.3ms, while at minimum, it must be at least 17ms."

There exists an ATX PSU standard, a guideline for all PSU OEMs, to make safe PSUs for PCs. ATX PSU standard is actually quite easy going and in reality, it should be far stricter. Still, there are PSU OEMs, who doesn't care much about ATX PSU standard and they don't put much effort into meeting the ATX PSU standard, when making PSUs. This results out of spec PSUs, which can damage the hardware they power.

Here's the latest iteration of ATX PSU standard, v3.0 and what it brought,
article: https://www.tomshardware.com/news/intel-atx-v3-psu-standard

Perhaps better example to make you understand the importance of ATX PSU standard;
Regarding cars, there exists a set of safety rules, that all cars must have certain safety features. Those include seat belts, ABS, air bags, crumple zones etc. Would you buy a car, who's manufacturer doesn't care about those safety rules and doesn't make their car as safe as other cars are? With less features, car is cheaper as well.
If not, then why not?

 

[kimbo_0331]

I will just add software tests like CPUID aren't accurate for testing a PSU on. Only real accurate test is a multimeter.

Event 41 (63) is reported by windows after an unexpected restart. It doesn't necessarily mean its the PSU. It can be, but its not an obvious sign. Windows used to mistakenly report one if you used fast startup.

@Aeacus lucky I even noticed this, not really here at the moment.

Colif,

Thanks for you input. I will certainly make sure that whomever helps to solve this dilemma receives the appropriate accolades. From what I am hearing from these peeps is it looks to be the PSU, or lack of quality therein.

Thanks,

 

[kimbo_0331]

Not all PSUs are created equal.

For example, can you tell me why your PSU costs easy $200 less than other 1.6kW PSUs,
pcpp: https://pcpartpicker.com/products/compare/LbtWGX,cJbwrH,tFvdnQ,X93gXL/


Sure, Seasonic and Corsair PSUs are 80+ Titanium (highest efficiency) but EVGA PSU is 80+ Platinum, which is only 2% more efficient over 80+ Gold. Yet, even the EVGA PSU costs 200 bucks more than your PSU.

Or another example,
pcpp: https://pcpartpicker.com/products/compare/RzfFf7,n34BD3/

Can you tell, why Seasonic PSU costs 50 bucks more than Enermax unit? Both are 650W, fully-modular and 80+ Gold efficiency. And which of the two would you get?



There exists an ATX PSU standard, a guideline for all PSU OEMs, to make safe PSUs for PCs. ATX PSU standard is actually quite easy going and in reality, it should be far stricter. Still, there are PSU OEMs, who doesn't care much about ATX PSU standard and they don't put much effort into meeting the ATX PSU standard, when making PSUs. This results out of spec PSUs, which can damage the hardware they power.

Here's the latest iteration of ATX PSU standard, v3.0 and what it brought,
article: https://www.tomshardware.com/news/intel-atx-v3-psu-standard

Perhaps better example to make you understand the importance of ATX PSU standard;
Regarding cars, there exists a set of safety rules, that all cars must have certain safety features. Those include seat belts, ABS, air bags, crumple zones etc. Would you buy a car, who's manufacturer doesn't care about those safety rules and doesn't make their car as safe as other cars are? With less features, car is cheaper as well.
If not, then why not?

Aeacus,

What PSU would you recommend for my build if I were to replace it? I don't need 1650w, but the though of perhaps future proofing my PC to upgrade to the GPU that perhaus needed two 16 pin connectors would have been compatible with my current setup. I should have paid a little more to the latest standards. I normally go with EVGA, but I want to utilize the latest ATX 3.0 standards with the 16pin connector socket for the RTX 4090. At this point just need a PSU that will provide sufficient standards for my setup, that I don't have to worry about.


-Intel Core i7-13700K 3.4 GHz 16-Core Processor
-NZXT Kraken Z73 RGB 52.44 CFM Liquid CPU Cooler
-MSI MPG Z790 EDGE WIFI ATX LGA1700 Motherboard
-OLOy Blade RGB 32 GB (4 x 8 GB) DDR5-5600 CL40 Memory
-Samsung 980 Pro 2 TB M.2-2280 PCIe 4.0 X4 NVME Solid State Drive
-Samsung 980 Pro 2 TB M.2-2280 PCIe 4.0 X4 NVME Solid State Drive
-MSI SUPRIM X GeForce RTX 4090 24 GB Video Card
-Thermaltake Toughpower GF3 TT Premium 1650 W 80+ Gold Certified Fully Modular ATX Power Supply
-Microsoft Windows 11 Pro OEM - DVD 64-bit
-Lian Li Uni Fan SL-Infinity 61.3 CFM 120 mm Fan
-Lian Li Uni Fan SL-Infinity 61.3 CFM 120 mm Fans 3-Pack
-Lian Li Uni Fan SL-Infinity 61.3 CFM 120 mm Fans 3-Pack
-Lian Li Uni Fan SL-Infinity 61.3 CFM 120 mm Fans 3-Pack
-Lian Li O11 Dynamic EVO ATX Mid Tower Case


If this fixes my problem, I will be sure to provide you with the credit. Thanks for your help.

Regards,

 

[Karadjgne]

Corsair was pretty much the innovator of putting the brakes on the gpu spikes by using a double OCP. There's one OCP with lower limits and short burst duration, just for the spikes, and the other higher limit long duration OCP for excessive power consumption.

Most of the other OEMs have adopted something similar in theory for several of their lines, but not all, since the 450w/550w class psus honestly don't see much benefit from such, gpus in that class don't really spike awful much.

The Corsair AXi 1500w psu was the single best psu on the market for several years, it's only been beaten by 1 psu and recently. The Corsair AXi 1600w. But for that psu, you'd be seriously stretching the budget at just over $600.

Imho, Corsair HX1200 or HX1500i, depending if you want iCue accessibility to customize outputs or not.

 

[Aeacus]

perhaps future proofing my PC

None of us can see into the future and can't tell if future GPUs require more than one 12VHPWR connector or not.
So, look at the situation at current moment. Unless you have the magical ability to see into the future.

What PSU would you recommend for my build if I were to replace it?

What i would suggest, if you want PSU with ATX 3.0 standard, would be Seasonic Vertex PX-1200. Though, PX-1200 is currently hard to find since there's RTX 40-series craze and everybody are after the beefy PSUs. Vertex GX-1200 is currently available,
Egg: https://www.newegg.com/seasonic-vertex-gx-1200/p/N82E16817320023

Personally, i'd go with Seasonic PRIME TX-1600 + Cablemod RT-series adapter (from 4x 8-pin to 1x 16-pin) and perhaps Cablemod angled adapter as well, for clean looks. Resulting in a highest efficiency (80+ Titanium), longest warranty (12 years) and best looks (color options with Cablemod).
Cablemod: https://store.cablemod.com/product-category/12vhpwr/

 

[JLW123178]

Hello,

Recently I just assembled a new gaming PC (See specs below). After about 3 weeks of gameplay, when I transitioned from a 2080ti to a 4090, I started having random reboots during gameplay only. Right now it has only occurred while playing Warhammer 3, and It Takes Two. At first I though it was software conflict because it was limited to one title, but after it started happening to a second game, I started to get worried all over again. I have ran Furmark (20 min) with no issues. I have played Warzone 2.0 with no issues (1 hr). I have played HL Alyx with no issues (2 hrs). I played Portal RTX with no issues (1 hr). It seems to be limited to both warhammer 3 (about 10 min) and It Takes Two for now (about 15 min). I will be purchasing A Plague Tale Requiem which from my research has also been a known source to duplicate this observation.

I have opened my event viewer to reviewer the Administrative Events at the time of the reboot. I get the Kernal0Power Event ID 41 ("the system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly). In the system information of the windows Logs at the time of the reboot it has Kernel Power Even ID 125 (ACPI thermal zone \_TZ.TZ00 has been enumerated).

I don't know what the problem could be. Right now I am leaning towards a faulty GPU or a faulty PSU. From the research I have conducted, others are having issues with the Thermaltake GF3 1650w PSU, but I don't want to complete an RMA without knowing for sure. To do this I ran the CPUID hardware monitor, but I am not tech savvy enough to discern any anomalies (see below for pictures). I can tell based on temps that it cannot be an overheating issue with the GPU, or the CPU which were my initial concerns. See below for those results after a recent unexpected reboot while playing warhammer 3.

If anyone is able to offer any insight onto a fix for this obersavtion, or at a minimum, provide some strong leads, I will issue a reward in form of a smile. After I have fixed the issue I will engage in celebratory gunfire and other festivities.

Things i've done to help troubleshoot based on what others have suggested:

• updated GPU drivers
• reintalled GPU drivers via Geforce Experience
• Used a DDU to erase all data pertaining to the 4090 and reinstalled drivers
• updated windows
• Turned off automatic restart
• Changed power options to balanced
• Changed power options to performance
• used a 2080ti (no random reboots)
• ran windows memory diagnostic
• disabled XMP
• enabled XMP
• ran furmark

System (Specs):
-Intel Core i7-13700K 3.4 GHz 16-Core Processor
-NZXT Kraken Z73 RGB 52.44 CFM Liquid CPU Cooler
-MSI MPG Z790 EDGE WIFI ATX LGA1700 Motherboard
-OLOy Blade RGB 32 GB (4 x 8 GB) DDR5-5600 CL40 Memory
-Samsung 980 Pro 2 TB M.2-2280 PCIe 4.0 X4 NVME Solid State Drive
-Samsung 980 Pro 2 TB M.2-2280 PCIe 4.0 X4 NVME Solid State Drive
-MSI SUPRIM X GeForce RTX 4090 24 GB Video Card
-Thermaltake Toughpower GF3 TT Premium 1650 W 80+ Gold Certified Fully Modular ATX Power Supply
-Microsoft Windows 11 Pro OEM - DVD 64-bit
-Lian Li Uni Fan SL-Infinity 61.3 CFM 120 mm Fan
-Lian Li Uni Fan SL-Infinity 61.3 CFM 120 mm Fans 3-Pack
-Lian Li Uni Fan SL-Infinity 61.3 CFM 120 mm Fans 3-Pack
-Lian Li Uni Fan SL-Infinity 61.3 CFM 120 mm Fans 3-Pack
-Lian Li O11 Dynamic EVO ATX Mid Tower Case

Thanks,

I had the same experience after I replaced my radeon 6700 xt with a 6950 xt. Everyone replied to my thread saying it's my PSU when its not even a year old and is 1000w which is more than enough. I ended up exchanging the GPU and I haven't had a kernel 41 error for the past two weeks. If you can try to exchange your GPU, you might have gotten unlucky like me with a bad gpu out of the box.

 

[--SID--]

Seems to be a promo for Seasonic, again.

https://pcpartpicker.com/product/RH...ed-fully-modular-atx-power-supply-pa-2g1bb-us is my recommandation because it's an outstanding unit. For a review: https://hwbusters.com/psus/nzxt-c1200-gold-psu-review-top-performance/

 

[kimbo_0331]

I had the same experience after I replaced my radeon 6700 xt with a 6950 xt. Everyone replied to my thread saying it's my PSU when its not even a year old and is 1000w which is more than enough. I ended up exchanging the GPU and I haven't had a kernel 41 error for the past two weeks. If you can try to exchange your GPU, you might have gotten unlucky like me with a bad gpu out of the box.

Estimable,

Thanks for your input. That will be the next step if the replacement PSU does not work. Hopefully not, but its certainly a possibility.

Thanks,