• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

5090 bought now having crashes

Knin3

Knin3

Knin3

Associate
Joined
30 Mar 2016
Posts
213
Hi guys,

I bought a 5090 msi, now im getting crashes from about 10 mins in upto about 30 mins. I thought it was the 5090 as I was fine up until this upgrade, previously had a 3090 but today I packed the 50 up for returns and put my 30 back in only to get the same crash. Ive now put the 50 back in and once again got a crash in outlast trials. This was the crash

Fatal error!

Unhandled Exception: EXCEPTION_ACCESS_VIOLATION writing address 0x0000000000000004

0x00007ff7cedaf62b TOTClient-Win64-Shipping.exe!TMapBase<FD3D12RayTracingShaderTable::FShaderRecordCacheKey,unsigned int,FDefaultSetAllocator,TDefaultMapHashableKeyFuncs<FD3D12RayTracingShaderTable::FShaderRecordCacheKey,unsigned int,0> >::Emplace<FD3D12RayTracingShaderTable::FShaderRecordCacheKey const &,unsigned int const &>() [D:\jenkins-steam-release-3-0\OPP\UnrealEngine4\Engine\Source\Runtime\Core\Public\Containers\Map.h:404]
0x00007ff7cedc3f34 TOTClient-Win64-Shipping.exe!SetRayTracingHitGroup() [D:\jenkins-steam-release-3-0\OPP\UnrealEngine4\Engine\Source\Runtime\D3D12RHI\Private\D3D12RayTracing.cpp:4000]
0x00007ff7cedc406b TOTClient-Win64-Shipping.exe!SetRayTracingHitGroupsTask() [D:\jenkins-steam-release-3-0\OPP\UnrealEngine4\Engine\Source\Runtime\D3D12RHI\Private\D3D12RayTracing.cpp:4031]
0x00007ff7cedb772e TOTClient-Win64-Shipping.exe!UE4Function_Private::TFunctionRefCaller<<lambda_87ed0d1645ee90dfb3a1eb37b987706e>,void __cdecl(void)>::Call() [D:\jenkins-steam-release-3-0\OPP\UnrealEngine4\Engine\Source\Runtime\Core\Public\Templates\Function.h:549]
0x00007ff7ccc41b02 TOTClient-Win64-Shipping.exe!TGraphTask<TFunctionGraphTaskImpl<void __cdecl(void),0> >::ExecuteTask() [D:\jenkins-steam-release-3-0\OPP\UnrealEngine4\Engine\Source\Runtime\Core\Public\Async\TaskGraphInterfaces.h:887]
0x00007ff7cc744046 TOTClient-Win64-Shipping.exe!FTaskThreadAnyThread::ProcessTasks() [D:\jenkins-steam-release-3-0\OPP\UnrealEngine4\Engine\Source\Runtime\Core\Private\Async\TaskGraph.cpp:1092]
0x00007ff7cc8f2e50 TOTClient-Win64-Shipping.exe!FTaskThreadAnyThread::ProcessTasksUntilQuit() [D:\jenkins-steam-release-3-0\OPP\UnrealEngine4\Engine\Source\Runtime\Core\Private\Async\TaskGraph.cpp:914]
0x00007ff7cc8f2e1c TOTClient-Win64-Shipping.exe!FTaskThreadBase::Run() [D:\jenkins-steam-release-3-0\OPP\UnrealEngine4\Engine\Source\Runtime\Core\Private\Async\TaskGraph.cpp:542]
0x00007ff7cc5396f4 TOTClient-Win64-Shipping.exe!FRunnableThreadWin::Run() [D:\jenkins-steam-release-3-0\OPP\UnrealEngine4\Engine\Source\Runtime\Core\Private\Windows\WindowsRunnableThread.cpp:86]
0x00007ff7cc53954a TOTClient-Win64-Shipping.exe!FRunnableThreadWin::GuardedRun() [D:\jenkins-steam-release-3-0\OPP\UnrealEngine4\Engine\Source\Runtime\Core\Private\Windows\WindowsRunnableThread.cpp:35]
0x00007ffc2122e8d7 KERNEL32.DLL!UnknownFunction []
0x00007ffc217914fc ntdll.dll!UnknownFunction []

Crash in runnable thread TaskGraphThreadNP 1

Can anyone make any sense of this and mayb help me, please :(
 
If you have any sort of system memory overclock then try removing it, that's usually the number one cause of crashes because everything runs in memory including the graphics card drivers. I would wager that a lot of times when people blame unstable graphics card drivers it's because their system memory is unstable.
 
mmj_uk said:
If you have any sort of system memory overclock then try removing it, that's usually the number one cause of crashes because everything runs in memory including the graphics card drivers. I would wager that a lot of times when people blame unstable graphics card drivers it's because their system memory is unstable.
Click to expand...
on my last system had to change the xpm to auto which weirdly said the same settings and it resolved the issue the crashes used to drive me round the bend
 
mmj_uk said:
I would wager that a lot of times when people blame unstable graphics card drivers it's because their system memory is unstable.
Click to expand...
I've seen some reports that the 5090 dumps a lot of extra heat on the memory, but not sure if that's specific to the FE?
 
Tetras said:
I've seen some reports that the 5090 dumps a lot of extra heat on the memory, but not sure if that's specific to the FE?
Click to expand...
Yeah makes sense but I think bog standard memory should be quite tolerant, the problem is most memory is factory overclocked/overvolted so close to its stable limit that they can't tolerate any extra heat.
 
Last edited:
Tetras said:
I've seen some reports that the 5090 dumps a lot of extra heat on the memory, but not sure if that's specific to the FE?
Click to expand...

I was about to disagree with this to protect my 5090's honor, and then I peered into my case.
Yea, that 5090FE fan is blasting nice warm air across the memory dimms :cry:

Thankfully it's not caused me any problems :)
 
Tetras said:
I've seen some reports that the 5090 dumps a lot of extra heat on the memory, but not sure if that's specific to the FE?
Click to expand...
3090 did it as well and yeah its something specific to the FE lord knows why they choose a one up/one down fan flow it means one of them always blasts hot air up all over the mobo/RAM it managed to kill a set of b-die Samsung ram I had it just cooked it had to replace it with more regular ram which seemed to put up with it, mind you I didn't have it that long after before I sold it
 
Last edited:
What RAM? Aggressive XMP/EXPO profiles require higher magnitude of voltage to run stable, which results in more heat generated. Above 1.4V it is recommended to seek more efficient cooling solution than the passive factory heatsink, especially when you got a high performance GPU blowing warm air at the DIMM sockets.
 
Blackjack Davy said:
3090 did it as well and yeah its something specific to the FE lord knows why they choose a one up/one down fan flow it means one of them always blasts hot air up all over the mobo/RAM it managed to kill a set of b-die Samsung ram I had it just cooked it had to replace it with more regular ram which seemed to put up with it, mind you I didn't have it that long after before I sold it
Click to expand...
one of those fans doesnt blow heat through the card-barrier, but i guess that changed for 5090 fe
 
brostradamus said:
one of those fans doesnt blow heat through the card-barrier, but i guess that changed for 5090 fe
Click to expand...
They also got rid of the holes on the cards back panel which allowed hot air out. So on the 5090 fe it has to go out into the case.

So yeah while the fe has always had this issue a bit, the 50 series has amplified it a bit.
 
Rawr said:
They also got rid of the holes on the cards back panel which allowed hot air out. So on the 5090 fe it has to go out into the case.

So yeah while the fe has always had this issue a bit, the 50 series has amplified it a bit.
Click to expand...

The hot air from the fan nearest the IO panel just blows up and then straight out of the rear case exhaust fan, as long as people have the rear case fan set to exhaust. I imagine most would.
So the main culprit for heating up the inside of the case is still the end fan, which on the 5090FE does indeed blow over the ram.

It's a good reason not to vertically mount the FE... It would blast all the hot air straight onto the motherboard and cook everything.
 
You must log in or register to reply here.
Back
Top Bottom