F5F Stay Refreshed Hardware Desktop The issue was resolved by unplugging the 24-pin connector.

The issue was resolved by unplugging the 24-pin connector.

The issue was resolved by unplugging the 24-pin connector.

C
cocothecat2
Member
66
10-17-2023, 11:45 AM
#1
I shared this on Saturday now, and it happened again. It didn’t go blue-screened; it just shut off and wouldn’t turn back on until I reconnected the 24-pin connector for both the CPU and GPU. They were around 50 to 60°C.
C
cocothecat2
10-17-2023, 11:45 AM #1

I shared this on Saturday now, and it happened again. It didn’t go blue-screened; it just shut off and wouldn’t turn back on until I reconnected the 24-pin connector for both the CPU and GPU. They were around 50 to 60°C.

D
dehunter456
Member
237
10-17-2023, 11:45 AM
#2
I'll attempt it once more if it crashes again.
D
dehunter456
10-17-2023, 11:45 AM #2

I'll attempt it once more if it crashes again.

N
nickel440
Member
141
10-17-2023, 11:45 AM
#3
To pinpoint the crash cause accurately, consider reviewing whocrashed data or checking event logs and sharing findings. Also, let me know if you're using the standard BIOS or an overclocked setup. From my experience with Ryzen 5000 systems, frequent crashes needing a full power cycle often point to VRM current limit issues. You might try adjusting the VRM settings in BIOS and run a stress test like Cinebench to observe changes.
N
nickel440
10-17-2023, 11:45 AM #3

To pinpoint the crash cause accurately, consider reviewing whocrashed data or checking event logs and sharing findings. Also, let me know if you're using the standard BIOS or an overclocked setup. From my experience with Ryzen 5000 systems, frequent crashes needing a full power cycle often point to VRM current limit issues. You might try adjusting the VRM settings in BIOS and run a stress test like Cinebench to observe changes.

A
agossie
Member
156
10-17-2023, 11:45 AM
#4
i'm using my 5900x on a 5ghz processor with ppt tdc and edc at full capacity. the msi x570 pro board is working fine, but it kept crashing again. after removing the front I/O and shorting the power pins, it stopped working without needing to unplug the 24-pin connector.
A
agossie
10-17-2023, 11:45 AM #4

i'm using my 5900x on a 5ghz processor with ppt tdc and edc at full capacity. the msi x570 pro board is working fine, but it kept crashing again. after removing the front I/O and shorting the power pins, it stopped working without needing to unplug the 24-pin connector.

F
Falymi
Member
113
10-17-2023, 11:45 AM
#5
I still suggest increasing the power ceiling of the motherboard. I mean not the PBO limits, because those instructions guide AMD’s power handling in a way that might not align with the VRMs’ capabilities. You should lift the current threshold so the CPU can utilize the VRMs without being cut off when the PBO algorithm signals sufficient power availability.
F
Falymi
10-17-2023, 11:45 AM #5

I still suggest increasing the power ceiling of the motherboard. I mean not the PBO limits, because those instructions guide AMD’s power handling in a way that might not align with the VRMs’ capabilities. You should lift the current threshold so the CPU can utilize the VRMs without being cut off when the PBO algorithm signals sufficient power availability.

E
EilaMummo
Member
116
10-17-2023, 11:45 AM
#6
I've reinitialized the BIOS and raised the power cap to 500W. The issue persists. I ran Cinebench for 20 minutes at 95% and it still crashed. Most recent failures occurred during gameplay of Honkai Star Rail.
E
EilaMummo
10-17-2023, 11:45 AM #6

I've reinitialized the BIOS and raised the power cap to 500W. The issue persists. I ran Cinebench for 20 minutes at 95% and it still crashed. Most recent failures occurred during gameplay of Honkai Star Rail.

1
13tj
Member
110
10-17-2023, 11:45 AM
#7
What kind of power source do you use? If it's under 1000 watts or doesn't come from a reputable PSU, there could be the well-known 3090 transient voltage problem where it draws excessive current briefly and activates overcurrent protection on the unit.
1
13tj
10-17-2023, 11:45 AM #7

What kind of power source do you use? If it's under 1000 watts or doesn't come from a reputable PSU, there could be the well-known 3090 transient voltage problem where it draws excessive current briefly and activates overcurrent protection on the unit.

S
superowl171
Junior Member
13
10-17-2023, 11:45 AM
#8
It's around 750 or 800 from Corsair. Acquired a new one about a year ago.
S
superowl171
10-17-2023, 11:45 AM #8

It's around 750 or 800 from Corsair. Acquired a new one about a year ago.

N
niclasdam
Member
175
10-17-2023, 11:45 AM
#9
N
niclasdam
10-17-2023, 11:45 AM #9

D
dazzlinglogan
Member
67
10-17-2023, 11:45 AM
#10
Loading screens appearing slow is typical. Often it's related to textures being compiled by the CPU, especially during the initial run. DX12 tends to stutter on first playthroughs because they don't use a shader cache and compile on the fly—a major issue until your second try. This should point to an SSD problem. Luckily we fixed the problem. If you want to maximize GPU performance, consider using a higher wattage power supply. Enjoy!
D
dazzlinglogan
10-17-2023, 11:45 AM #10

Loading screens appearing slow is typical. Often it's related to textures being compiled by the CPU, especially during the initial run. DX12 tends to stutter on first playthroughs because they don't use a shader cache and compile on the fly—a major issue until your second try. This should point to an SSD problem. Luckily we fixed the problem. If you want to maximize GPU performance, consider using a higher wattage power supply. Enjoy!