Alle diskussioner > Steam-fora > Hardware and Operating Systems > Trådoplysninger
LionGoodWay 12. nov. kl. 14:39
VRam
Good time of day. An error appears in the event log "The description for the event ID 153 from the nvlddmkm source cannot be found. The component causing this event is not installed on this local computer or is corrupted. Install or restore the component on the local computer.
\Device\Video3
Error occurred on GPUID: 100"
After successfully passing the VRAM video memory test in OCCT, at any test duration, and at any percentage load of video memory. During Extreme stress tests, the OSST program itself does not register errors, the test is also successful and no such error is recorded in the event log, there are no artifacts, BSOD. According to the advice from the forums who encountered a similar error, I disabled the Nvidia sound drivers in the device manager, but the error reappears after successfully passing the Vram test, the OCCT program itself declares passing the test without errors.
< >
Viser 1-10 af 10 kommentarer
smallcat 12. nov. kl. 14:45 
what could help - clean install of NVidia drivers, Windows reset with cloud image download . Check the SSD health .it might be failing .benchmark it and see the results .Undo any overclocking
Sidst redigeret af smallcat; 12. nov. kl. 14:46
#1
LionGoodWay 13. nov. kl. 1:37 
Oprindeligt skrevet af smallcat:
what could help - clean install of NVidia drivers, Windows reset with cloud image download . Check the SSD health .it might be failing .benchmark it and see the results .Undo any overclocking
So it doesn't mean that the VRAM is damaged? Is she okay?
#2
ArielNugie 13. nov. kl. 1:50 
It's most likely just a driver-side hiccup (nvlddmkm/TDR) rather than a real VRAM failure. Windows logs that event when the NVIDIA driver hangs for a moment and doesn't respond fast enough, but OCCT or other stress tests won't always catch that because they don't trigger the same code path the driver uses during normal operation.

If the card passes heavy stress tests, no artifacts, no crashes, no throttling, then it's usually just a software/driver quirk rather than dying hardware. But if the event keeps appearing after a clean driver install, it might be worth testing the GPU in another system to rule things out.
Sidst redigeret af ArielNugie; 13. nov. kl. 1:52
#3
LionGoodWay 13. nov. kl. 2:24 
Oprindeligt skrevet af ArielNugie:
It's most likely just a driver-side hiccup (nvlddmkm/TDR) rather than a real VRAM failure. Windows logs that event when the NVIDIA driver hangs for a moment and doesn't respond fast enough, but OCCT or other stress tests won't always catch that because they don't trigger the same code path the driver uses during normal operation.

If the card passes heavy stress tests, no artifacts, no crashes, no throttling, then it's usually just a software/driver quirk rather than dying hardware. But if the event keeps appearing after a clean driver install, it might be worth testing the GPU in another system to rule things out.
This error is registered once every other time exclusively after the VRAM test in OCCT, there is no such error in games and stress tests.
#4
smallcat 13. nov. kl. 7:17 
no artifacts , crashes ,perf problems; VRAM stress test passes cleanly . So the VRAM is OK .it s a software issue
Sidst redigeret af smallcat; 13. nov. kl. 7:17
#5
LionGoodWay 13. nov. kl. 7:43 
Oprindeligt skrevet af smallcat:
no artifacts , crashes ,perf problems; VRAM stress test passes cleanly . So the VRAM is OK .it s a software issue
Thanks
#6
HeyYou 13. nov. kl. 7:46 
Do a clean uninstall of the existing drivers. (download the new ones first....) Reboot, install new drivers. Run something like Furmark for 15 to 20 minutes. If it doesn't choke and die, you are good to go.
#7
LionGoodWay 13. nov. kl. 7:52 
Oprindeligt skrevet af HeyYou:
Do a clean uninstall of the existing drivers. (download the new ones first....) Reboot, install new drivers. Run something like Furmark for 15 to 20 minutes. If it doesn't choke and die, you are good to go.
The existing driver was installed through a clean install, checked the box in the Nvidia installer, tests were conducted in Furmark for artifacts, and a stress test was performed in OSST. There are no driver failures in games, as well as in stress tests. It's just that for some reason he gets cranky every other time, in the VRAM test, when the test is successful, and even if you interrupt it for the sake of interest, error 153 will also be called in the event log. I think everything is great, there is no need to force the thermal interface with eternal tests...
#8
HeyYou 13. nov. kl. 7:54 
If it works just fine in games, ignore the errors, and call it a day. If it ain't broke, don't fix it.
#9
LionGoodWay 13. nov. kl. 7:57 
Oprindeligt skrevet af HeyYou:
If it works just fine in games, ignore the errors, and call it a day. If it ain't broke, don't fix it.
Well, I'll demolish the driver for the sake of interest and install it again, let's see what happens. ((the error returned after canceling/successfully passing the VRAM test, despite demolition via DDU xD))
Sidst redigeret af LionGoodWay; 13. nov. kl. 8:14
#10
< >
Viser 1-10 af 10 kommentarer
Per side: 1530 50

Alle diskussioner > Steam-fora > Hardware and Operating Systems > Trådoplysninger
Dato opslået: 12. nov. kl. 14:39
Indlæg: 10
Start en ny diskussion

Diskussionsregler og retningslinjer