I’ve got an MSI Suprim RTX 5080 paired with an MSI Z690 Carbon motherboard, and I’m running into a strange issue: my GPU is drawing 140W of power even when it’s completely idle - no load, nothing running at all. I’ve confirmed this with MSI Afterburner, as well as GPU-Z and HWInfo.
Here are the details:
GPU State: P8 (idle)
Temperature: 35°C
Power Consumption: 140W (even at idle)
I’ve tested this on both Windows 11 (clean install) and Linux, and the results are the same. I’ve also tried reinstalling the Nvidia drivers (tried both the latest and an older version) and resetting the UEFI settings to defaults, but the issue persists.
Additionally, I measured the total power consumption of the entire PC while idle, and it’s around 55W, so it seems like the GPU is reporting incorrect power consumption.
Has anyone experienced something similar? Any suggestions on how to fix this or what might be causing it? Appreciate any help!
@cats are you still having problems with 576.52? I am also seeing this strange power consumption with my 4090. I have a HX1500i PSU which has power monitoring and how is the PSU reporting 120w consumption with the GPU using 115w? I switched from AM4 to AM5 and now I’m seeing these readings and I don’t think it’s possible. I also see about 40W more power during gaming.
hey @lucas.casul, I wasn’t able to resolve this issue, so I sent it back to the service center a week ago. They said it will take a few weeks to diagnose and fix the issue. Will share an update later.
@cats But your card has had other issue? Mine works fine, no problems, including temperature, just this sudden change in power consumption that doesn’t seem accurate
I experienced the same behaviour. When i first bought the card its was showing correct values on power consumption.
However after a few weeks it started to show higher values than the actual consumption.
The card worked OK, but I found that i can’t overclock it and use higher GPU frequency.
The card couldn’t keep even 3Ghz because it was hitting the pawer draw limit - showed 360W while it was ~280.
@cats I tried with other drivers, flashing the gpu and changing the cable. Nothing worked, so I reverted the bios, but now it’s getting worse. Heavily throttling due to power limiting, but temperatures are ok (gpu - 58c, vrm 64c, memory 70c, ambient 29c). I will try a deep clean to see if things change. I’m starting to think that something went wrong with the power delivery, VRM or something and I don’t think it’s the PSU. This is annoying 😥. Anyway, I hope they fix your card or get a brand new one ASAP.
There are multiple possible reasons for this issue.
It could be a hardware problem with the shunt resistor, which is used to measure the voltage drop and estimate power consumption. The IC chip responsible for measurement calculations could also be affected.
Voltage instability might be the cause this, for example, due to a loose cable connection.
There is also a chance that the value is being incorrectly interpreted by the driver or firmware.
If your card is still under warranty, try contacting the shop or vendor.
@lucas.casul, so after 2 weeks in service, MSI concluded that the issue can’t be fixed quickly and my local shop returned my money since there is no available replacement at the moment. They didn’t share any details …