[SOLVED] Games crash with 0xc0000005

Apr 30, 2021
36
0
30
0
I've overclocked/undervolted my CPU and I've run several stress tests for hours and it was stable, but I noticed that some games started to crash with error 0xc0000005.
The event viewer has no WHEA errors, so I don't know which core crashes or any hardware failure.

Can it be related to the CPU OC/UV? I also have an XMP profile activated, but I feel that this happened since I tweak the CPU.
Can games crash like that due to OC while stressing it with hours of stress tests passed?
 
Sorry, my spec is:
  • CPU - AMD Ryzen 9 5900x
  • Motherboard - Asus Strix X570-F Gaming
  • Memory - Corsair Vengeance LPX 64 GB (2 x 32 GB) 3200 CL16
  • Cooler - be quiet! Dark Rock Pro 4
  • Power supply - Antec HCG650 Gold
I use the PBO Curve Optimizer and undervolted all cores by 25 (with reducing the PPT to 110 to limit the overclocking). The reason that I undervolt is to reduce the CPU temrature a little bit more, since while gaming it can raise up to 85C (stable), which is OK, but I prefer to keep it under 80C.

BTW, is there any way to know which core has crashed instead of raising the voltage in all cores? Except of this error, I can't see other errors in the Event Viewer.
Try reducing the Curve Optimizer undervolt setting to get it stable.

Also, instead of reducing the PPT limit, which lowers maximum processor performance under any condition even when temps are well below 80C, set the platform thermal limit to 80C, your desired limit. That will let the processor hit the fuzed power limit for short bursts whenever temps are below the 80C limit, a better setup especially for gaming.

That's a pretty good cooler...how's your case set up? If it's not well ventilated, with case fans ramping up as the GPU does, all it ends up doing is heating the CPU with the GPU's enormous heat output. I'm assuming your GPU compares well to the rest of your system specs, of course, and any modern one can double down as your room heater. I'd really expect temps in the mid-upper 70's while gaming unless running a fixed, manual overclock.
 
Last edited:

iPeekYou

Honorable
Jul 7, 2014
322
57
10,790
8
I've overclocked/undervolted my CPU and I've run several stress tests for hours and it was stable, but I noticed that some games started to crash with error 0xc0000005.
The event viewer has no WHEA errors, so I don't know which core crashes or any hardware failure.

Can it be related to the CPU OC/UV? I also have an XMP profile activated, but I feel that this happened since I tweak the CPU.
Can games crash like that due to OC while stressing it with hours of stress tests passed?
Sure can, I had OC/UV settings that passed OCCT both small and large data set tests for an hour each, only to fail in AC:Odyssey within 10 minutes or so.
 

Jason H.

Distinguished
I've overclocked/undervolted my CPU and I've run several stress tests for hours and it was stable, but I noticed that some games started to crash with error 0xc0000005.
The event viewer has no WHEA errors, so I don't know which core crashes or any hardware failure.

Can it be related to the CPU OC/UV? I also have an XMP profile activated, but I feel that this happened since I tweak the CPU.
Can games crash like that due to OC while stressing it with hours of stress tests passed?
Firstly, when posting to this forum you need to supply your full system specs. Including your overclocks and manual voltages.

With that said, Yes. The issue with stress tests is they are synthetic loads. Its not a end all be all. While gaming, much more is going on, textures being loaded in and out, reading/writing from the HDD/SSD, reading and writing from the ram cache, cpu to gpu data being processed... and much more.

Does the issue still happen without the OC? If not, then yes absolutely it is the OC. I would maybe try raising the voltage a bit.. how much have you undervolted it? I cant give you much advice without knowing your specs.
 
Apr 30, 2021
36
0
30
0
Firstly, when posting to this forum you need to supply your full system specs. Including your overclocks and manual voltages.

With that said, Yes. The issue with stress tests is they are synthetic loads. Its not a end all be all. While gaming, much more is going on, textures being loaded in and out, reading/writing from the HDD/SSD, reading and writing from the ram cache, cpu to gpu data being processed... and much more.

Does the issue still happen without the OC? If not, then yes absolutely it is the OC. I would maybe try raising the voltage a bit.. how much have you undervolted it? I cant give you much advice without knowing your specs.
Sorry, my spec is:
  • CPU - AMD Ryzen 9 5900x
  • Motherboard - Asus Strix X570-F Gaming
  • Memory - Corsair Vengeance LPX 64 GB (2 x 32 GB) 3200 CL16
  • Cooler - be quiet! Dark Rock Pro 4
  • Power supply - Antec HCG650 Gold
I use the PBO Curve Optimizer and undervolted all cores by 25 (with reducing the PPT to 110 to limit the overclocking). The reason that I undervolt is to reduce the CPU temrature a little bit more, since while gaming it can raise up to 85C (stable), which is OK, but I prefer to keep it under 80C.

BTW, is there any way to know which core has crashed instead of raising the voltage in all cores? Except of this error, I can't see other errors in the Event Viewer.
 
Sorry, my spec is:
  • CPU - AMD Ryzen 9 5900x
  • Motherboard - Asus Strix X570-F Gaming
  • Memory - Corsair Vengeance LPX 64 GB (2 x 32 GB) 3200 CL16
  • Cooler - be quiet! Dark Rock Pro 4
  • Power supply - Antec HCG650 Gold
I use the PBO Curve Optimizer and undervolted all cores by 25 (with reducing the PPT to 110 to limit the overclocking). The reason that I undervolt is to reduce the CPU temrature a little bit more, since while gaming it can raise up to 85C (stable), which is OK, but I prefer to keep it under 80C.

BTW, is there any way to know which core has crashed instead of raising the voltage in all cores? Except of this error, I can't see other errors in the Event Viewer.
Try reducing the Curve Optimizer undervolt setting to get it stable.

Also, instead of reducing the PPT limit, which lowers maximum processor performance under any condition even when temps are well below 80C, set the platform thermal limit to 80C, your desired limit. That will let the processor hit the fuzed power limit for short bursts whenever temps are below the 80C limit, a better setup especially for gaming.

That's a pretty good cooler...how's your case set up? If it's not well ventilated, with case fans ramping up as the GPU does, all it ends up doing is heating the CPU with the GPU's enormous heat output. I'm assuming your GPU compares well to the rest of your system specs, of course, and any modern one can double down as your room heater. I'd really expect temps in the mid-upper 70's while gaming unless running a fixed, manual overclock.
 
Last edited:
Apr 30, 2021
36
0
30
0
Try reducing the Curve Optimizer undervolt setting to get it stable.

Also, instead of reducing the PPT limit, which lowers maximum processor performance under any condition even when temps are well below 80C, set the platform thermal limit to 80C, your desired limit. That will let the processor hit the fuzed power limit for short bursts whenever temps are below the 80C limit, a better setup especially for gaming.

That's a pretty good cooler...how's your case set up? If it's not well ventilated, with case fans ramping up as the GPU does, all it ends up doing is heating the CPU with the GPU's enormous heat output. I'm assuming your GPU compares well to the rest of your system specs, of course, and any modern one can double down as your room heater. I'd really expect temps in the mid-upper 70's while gaming unless running a fixed, manual overclock.
My GPU is Gigabyte RTX 2070 Windforce and my case is Antec P110 Silent.

While Cinebench, the CPU temperature is around 68-73C. Gaming raises the temperature, and I believe that is related to the GPU heat that rises up to the CPU cooler. My case probably not so ventilated since it is a silent one, but all the fans are installed. BTW, it does not reach 85C in every game, just in some...

And I'm not looking for overclocking. The stock CPU's performance is excellent. I've lowered the PPT to limit the OC that the CPU does by itself since it has more room to do so due to the UV. In my tests, I've seen high correlation between the power consumption and heat - high PPT = high temperature. I didn't see cases where the PPT limits the processor performance when it was cool :D.

Rasing the PPT again will cause the CPU to OC and raise it voltage. higher clockspeed is much higher voltage (exponential ratio), so I don't want to let it get high voltage even if it cool, to extends its life as much as possible.
 
....

Rasing the PPT again will cause the CPU to OC and raise it voltage. higher clockspeed is much higher voltage (exponential ratio), so I don't want to let it get high voltage even if it cool, to extends its life as much as possible.
Well...two things...one is try Eco-Mode settings. It does a much better job of effectively lowering your processor's TDP so it will run cooler and does it coherently with other settings so it remains stable.

But two, and the most important: just leave it alone. AMD engineers have worked an awful lot of long hours on this processor so it functions well enough with default settings. They've bet a multi-billion dollar company's financial future on it, in fact, by selling millions of these things set up to work this way. So trust them and leave it in default settings, it will last at least 30 years which is way, way, WAY beyond the point where it has any level of computing relevancy in anyone's life. By then we'll have iRings that project 3D holographic games in total immersion with 5-sense synaptic feedback and human-level AI NPC's by then.
 
Last edited:
Apr 30, 2021
36
0
30
0
Well...two things...one is try Eco-Mode settings. It does a much better job of effectively lowering your processor's TDP so it will run cooler and does it coherently with other settings so it remains stable.

But two, and the most important: just leave it alone. AMD engineers have worked an awful lot of long hours on this processor so it functions well enough with default settings. They've bet a multi-billion dollar company's financial future on it, in fact, by selling millions of these things set up to work this way. So trust them and leave it in default settings, it will last at least 30 years which is way, way, WAY beyond the point where it has any level of computing relevancy in anyone's life. By then we'll have iRings that project 3D holographic games in total immersion with 5-sense synaptic feedback and human-level AI NPC's by then.
That's true, but maybe my cooling system is not efficient enough. It can reach the heat limit and I want to add more margin. Moreover, less voltage is always a good thing.

The ECO mode is too low, it reduces the CPU performance much more than UV. I reduced the UV to 20 instead of 25, and the games are much more stables now. And with that, the CPU temp is stable under 80C without losing too much performance.
 

ASK THE COMMUNITY

TRENDING THREADS