BSOD when hot temperatures

[Nasko]

Hi all,

I've recently (in July) ordered a gaming computer here, and received it perfectly fine, it worked really well.

But when the high temperatures started to hit my city (around 30°C), I started to get some BSOD : "system service exception".
And this only occurs on high temperatures outside, on big games, never on cool ones (20°C is perfectly fine), after few minutes/hours playing.

So that's now 100% sure the BSOD I'm getting are caused by an overheating component ... but which one, and why?

I've then checked my CPU, an AMD 7 5800x which is watercooled by Corsair Hydro 100X, but temperatures were fine to me, CPU temp never displays me anything above 75°C, even before crashes.

GPU is cool too, my RTX 3070 EVGA triple fan never exceeds 65°C (at worse).
I've tested my RAM on the full MemTest test from the BIOS, everything was fine too.

And logs quite don't help me to know what is causing those BSODs.

Here are my logs from eventviewer, system info and drivers through WeTransfer links :

System Event viewer : https://we.tl/t-lICfH4kX2u
Application Event viewer : https://we.tl/t-Q04oY4Fv7B
System Info : https://we.tl/t-zFFb38cUxd
Drivers : https://we.tl/t-AYiWeLM1dI


If you can help me in any way!

Thanks all,
Nasko


My full spec from this website (sorry for bad translations, it's google) :

Case
CASE MIDDLE TOWER PCS ARGB P209
Processor (CPU)
CPU Eight-Core AMD Ryzen 7 5800X (3,8GHz-4,7GHz / 36Mo CACHE / AM4)
Motherboard
ASUS ® X570-TUF MORE GAMING (USB 3.2 Gen 2, PCIe 4.0 , CrossFireX) - ARGB compatible!
Memory (RAM)
16 GB Corsair VENGEANCE DDR4 3200 MHz (2 x 8 GB)
Graphics card
8 GB NVIDIA GEFORCE RTX 3070 - HDMI, DP, LHR
1 st storage disk
1TB Seagate Barracuda SATA-III 3.5 inch HDD 6 GB / s, 7200 T / MIN 64MB CACHE
1 st disk SSD M.2
1TB PCIe SSD M.2 Intel® 670P NVMe (up to 3500 MB / s R | 2500 MB / sW)
Food
CORSAIR 650 W TXm SERIES ™ SEMI-MODULAR 80 PLUS® GOLD, ULTRA-QUIET
Power
cable 1 European power cable, 1 m (C13 / 14)
Processor
cooling CPU cooler Corsair H100x Hydro Series high performance
Thermal paste
PASTE STANDARD THERMAL FOR EFFICIENT COOLING HIGH DEF AUDIO
sound card
(STANDARD) 6 CHANNELS (5.1) INTEGRATED
Network card
PORT 10/100/1000 GIGABIT LAN
Wireless network
card PCI-E CARD WI-FI 802.11N 300 Mbps / 2, 4 GHz
USB / Thunderbolt options
MIN. 2 x USB 3.0 PORTS and 2 x USB 2.0 PORTS @ REAR PANEL + MIN. 2 FRONT PORTS
Operating System
OPERATING SYSTEM NOT REQUIRED
Operating System Language
France / French Republic - French
Windows Recovery Media
NOT REQUIRED RECOVERY MEDIA
Office software
FREE Microsoft 365® trial (operating system required) for 30 days
Antivirus
NO ANTIVIRUS Microsoft® Edge
browser
(Windows 10 only)
Warranty
Warranty gold 3 years (2 years collection and return, 2 years parts, 3 years labor)
Delivery
2 DAY DELIVERY iN METROPOLITAN FRANCE
Construction time
standard assembly: within the limits of available stocks of the products in pre-order
Welcome
booklet PCSpecialist welcome booklet - France
Logo marking
PCSpecialist Logo


EDIT : Updated the file names

 

[Martinr36]

When you're seeking help with a BSOD

Don't Panic! Many people panic when they get a BOSD, partly this is because they happen unexpectedly and because they usually happen without warning, but also because many believe they mean you have a hardware problem. The good news is that in the majority of cases a BSOD is a software problem...

www.pcspecialist.co.uk

 

[ubuysa]

First off, your Application log is filled with errors for a Brave browser extension for something called Greaselion. I'd be surprised if that caused a BSOD but it's not impossible.

There are also numerous and repeating entries for for BraveUpdateHelper.msi. There's no indication of an error but clearly something within Brave isn't happy.

There are also numerous and repeating reconfiguration messages for Google Update Helper. Quite what that means I have no idea, but they repeat regularly.

The System log contains several WHEA (Windows Hardware Error Architecture) entries for a machine check exception. One is a 'cache hierarchy error' and the other is a 'Bus/Interconnect error'. These are typical of the BSODs we were seeing a while back with an AMD build and 4 sticks of 3600MHz RAM.

I see you only have two sticks of 3200MHz RAM so it may not be the same issue. Nevertheless I suggest you phone PCS and ask whether there is a BIOS/AGESA update for your build. That (I believe) was the solution for these earlier (timing?) errors.

If you think it's temperature related then these errors would point at either RAM or CPU...

 

[Scott]

Shot in the dark here but 650w seems seriously low for your system. Wouldn't surprise me in the slightest if the temperatures lead to inefficiencies with the PSU and shot a power limit BSOD.

The Case and PSU would not have been recommended on this build.

 

[Nasko]

Shot in the dark here but 650w seems seriously low for your system. Wouldn't surprise me in the slightest if the temperatures lead to inefficiencies with the PSU and shot a power limit BSOD.

The Case and PSU would not have been recommended on this build.

If the PSU was the problem, I assume that the PC would just crash without any BSOD, moreover I can see that on high temperatures, and with big games running, I have a power usage of about 500W average (clamp meter). Moreover, PCS told me when I built the PC to downgrade my 750W PSU to a 650W one, because even with a 20% margin, my PC specs couldn't consume that much power.

About the case, it is actually kinda hot when my PC's crashing, BUT none of my components in the inside of the case are actually too hot, as I mentioned above. My thoughts are that my case is probably the issue, but I have no idea what is the overheating component.

First off, your Application log is filled with errors for a Brave browser extension for something called Greaselion. I'd be surprised if that caused a BSOD but it's not impossible.

There are also numerous and repeating entries for for BraveUpdateHelper.msi. There's no indication of an error but clearly something within Brave isn't happy.

There are also numerous and repeating reconfiguration messages for Google Update Helper. Quite what that means I have no idea, but they repeat regularly.

The System log contains several WHEA (Windows Hardware Error Architecture) entries for a machine check exception. One is a 'cache hierarchy error' and the other is a 'Bus/Interconnect error'. These are typical of the BSODs we were seeing a while back with an AMD build and 4 sticks of 3600MHz RAM.

I see you only have two sticks of 3200MHz RAM so it may not be the same issue. Nevertheless I suggest you phone PCS and ask whether there is a BIOS/AGESA update for your build. That (I believe) was the solution for these earlier (timing?) errors.

If you think it's temperature related then these errors would point at either RAM or CPU...

I'm gonna check that out, and I updated my BIOS few weeks ago + downclocked my RAM, funny part is that at same temperatures I had significantly less BSOD, but I still have some.
I'm also gonna clean that Brave's mess.


Thanks all !

 

[Scott]

If the PSU was the problem, I assume that the PC would just crash without any BSOD, moreover I can see that on high temperatures, and with big games running, I have a power usage of about 500W average (clamp meter). Moreover, PCS told me when I built the PC to downgrade my 750W PSU to a 650W one, because even with a 20% margin, my PC specs couldn't consume that much power.

About the case, it is actually kinda hot when my PC's crashing, BUT none of my components in the inside of the case are actually too hot, as I mentioned above. My thoughts are that my case is probably the issue, but I have no idea what is the overheating component.

Never assume

If it was an all out fail it would just reset, I don't mean an all out fail. I mean at the limit of it's performance ability while being stressed by temperature. This could cause what we would call an un-clean voltage line which could pass as a blip in the process. When this happens, an instruction or a clock gets missed and this causes a process failure..... if strong enough would cause a BSOD.

I take it you are checking the power usage at the socket side? 500W average using a clamp ammeter and then calculating isn't really going to tell you anything. You would need a through power meter to get anything accurate. You also need to consider the efficiency of the PSU and the fact that the efficiency is rated, and limited, by temperature.

If you are referring to the configurator stating the power limitations, I think everyone would agree that they are no longer accurate. They take very basic understandings of the power utilisations of both the CPU and the GPU into the equation. IIRC they tend to use TDP ratings rather than sustained upper limits. With the auto-overclocking nature of modern day CPUs and GPUs you can far exceed the TDP ratings of the core parts. NVidia themselves recommend a 750w PSU for the 3070.... as an example... and the 5800X is no slouch at full tilt either.

I doubt you will find anything specific, certainly not from the monitoring software. I don't think the system is fit for the purpose of its intended use and it would have never been recommended to proceed in such a way by any enthusiast on here.

But hey, what do I know