Board power draw exceeding whopping 550w

[FATAHALI00]

Okay so, i bought a dell m6800 workstation laptop having specs: i7 4900mq, 8 gigs ram, 256 ssd, 512 hdd and a 4 gb nvidia quadro k3100m graphics card. Through it is not a laptop that would run high end modern games but it did the job for me as a student of software engineering and playing simple games in my free time, i ran gta 5 on high setting which was extremely smooth 1920x1080 resolution, moreover i played mortal kombat xl with high graphics and no lag either, other games included, watch dogs, tekken 7, etc etc.
Now, one day i turned on my laptop and all these games started to lag at the point of 2 fps almost i got worried but upon cheling the graphics card showed no issues but still wouldnt run anything. Upon looking at the gpuz screen the graphics card was stuck at 162 memory clock and 135 core clock upon no matter what i did, the perf cap showed "PWR" and was a long green line in that column and i noticed my total power draw was well over exceeding 500w so was the pcie slot draw, i reinstalled windows multiple times, reinstalled drivers, reflashed the bios, tried to put the laptop at the maximum Performance it can provide but nothing improved the graphics cards clock speeds
If someone can tell me what the issue is that would be really helpfull

 

[ShrimpBrime]

It's a 75w card, but from what you describe, a mosfet might have burnt. (just a guess) Send it in to the repair shop, it's likely a physical defect of some kind.

 

[Toothless]

Bad sensor. Really nothing wrong with it.

 

[GerKNG]

your motherboard would be dead if that would be even close to real.

if any software sensor shows anything beyond +/- 20% of what's normal it's pretty safe to say it was a readout error.

 

[Redwoodz]

I would say a likely faulty GPU, if it was just bad sensor you would have the same performance. Maybe install HWinfo for more info on what the whole system is doing.

 

[R-T-B]

if it was just bad sensor you would have the same performance.

I mean gpus are watt limited based on the sensor readout no?

 

[DirtyDingusMcgee]

Something fishy. The a/c adapter couldn't supply anywhere close to that much power, and would die or catch fire trying. It came with a 240w charger, which is pretty beefy, but it isn't gonna pull 550w+ for long or at all.

https://i.dell.com/sites/doccontent...Documents/Dell_Precision_M6800_Spec_Sheet.pdf

 

[Vincero]

I'd trust the clock speed readings.... not so much the power readings - I think they are BS numbers.
More than 500W being pulled in a laptop on the GPU is highly unlikely - even using Dell's 20V adapter standard, that would need to be supplied with over 25 amps - they do not supply that much.

Have you actually checked / blown out the fan+heatsink on it? It could be throttling....
Also, does the laptop use the Intel+Nvidia graphics switching (if it's available)? Have you switched it to just use the Nvidia output only?

 

[Ruru]

Check what HWINFO64 says.

 

[DR4G00N]

It's a fault with the power sensing circuit, it's not too rare of a problem. A repair shop might be able to fix it but more than likely you'll have to replace/upgrade it.

 

[Toothless]

Check what HWINFO64 says.

Won't matter if the sensor is bad.

 

[Ruru]

Won't matter if the sensor is bad.

Yeah, thought just that maybe GPU-Z misreads it and some other software doesn't. Seen that happen from time to time.

 

[Dr. Dro]

I mean gpus are watt limited based on the sensor readout no?

It does. If performance is affected, then it's a damaged shunt.

 

[Ruru]

It does. If performance is affected, then it's a damaged shunt.

Didn't a shunt failure cause a GPU to underclock massively?

 

[R-T-B]

Didn't a shunt failure cause a GPU to underclock massively?

Is that not whats happening here though?

 

[Toothless]

Is that not whats happening here though?

Yeah was gonna say.. a quick search on the card shows it's clocking just fine.

 

[FATAHALI00]

Bad sensor. Really nothing wrong with it.

But the clock speeds of the card are having the issue aswell if only the sensor was bad, then would it not be the readings were bad and the performance would be similar to what i had before rather than the extreme frame drops and the frozen clock speeds

 

[W1zzard]

But the clock speeds of the card are having the issue aswell if only the sensor was bad, then would it not be the readings were bad and the performance would be similar to what i had before rather than the extreme frame drops and the frozen clock speeds

What's happening is that due the faulty sensor, the GPU thinks it's drawing too much power, and throttles down, in your case to minimal clocks.

 

[Naki]

I have the very same specs workstation laptop, but from HP!

(Obviously, not 8 GB RAM as that would not work for me due to many VMs I have - was 16 GB originally, then upped to 32 GB (the max). My storage/disks are also much bigger, of course.)

The CPU & GPU are identical though!

Do you guys want me to test something and send screenshots from my side?
(Using Windows 11 OS on the laptop for a few years, was on Windows 10 prior.)