Error occurs during stress on both CPU and GPU simultaneously, without affecting just one component.
Error occurs during stress on both CPU and GPU simultaneously, without affecting just one component.
Hello, Your system frequently shuts down when both the CPU and GPU are under heavy use, even though each component works fine individually. This leads you to think the power supply unit might be the issue. I’m seeking additional perspectives before purchasing a new one. You mentioned difficulty generating dump files due to errors during creation; the crashes occur with Event ID 41, originating from Kernel-Power. The dump file error ID is 161, linked to volmgr. Any guidance would be greatly appreciated, and I’m ready to provide any details you have. Your specs are: GPU RTX 3080, CPU Ryzen 7 1700X, 16GB Corsair Vengeance DDR4 RAM, ASRock x370 Killer, and a Corsair RM 750x PSU.
Consider lowering the power consumption of the CPU and GPU to assess whether a lighter workload prevents the crash. Turn off SMT on the CPU and adjust the GPU clock speed using MSI Afterburner or your preferred utility.
The RTX 3080 is powered using a single daisy-chained PCI-E 8-pin cable from the power supply unit, or with two distinct cables from the same PSU.
This approach appears to be working well, increasing GeForce power to 115% for the 3080 leads while causing some instability. Lowering it to 90% seems stable enough for Prime95 and Furmark. It looks like you're coordinating multiple connections now. Do you need guidance on using two separate cables?
When two separate cables are connected and issues persist, contact Corsair for assistance—their power unit is likely sufficient for this configuration.