[SOLVED] GPU Fan Spinning Up To 60000000 RPM?

[bigjames489]

Hi all,

I have an Asus ROG STRIX GTX 1660 Ti graphics card. It's been working fine for a good year and a half now. However, today I could audibly hear the GPU fans ramping up and down. This only seems to happen when idle and not on a game.

Here are my full system specs:

CPU: Ryzen 7 3800XT - Overclocked to stable 4.4 GHz at 1.1V
Motherboard: ASUS ROG Crosshair VI Hero (Wi-Fi AC)
CPU Cooler: MSI MAG CORELIQUID 240R
GPU: Asus ROG STRIX GTX 1660 Ti
SSD: Samsung 970 EBO 500GB
SSD: SanDisk SSD PLUS 1TB
PSU: Corsair Vengeance 750M (750 Watts, Year and a half old, in working and good condition)
RAM: CRUCIAL Ballistic 2x8GB 3200 MHz
Case: Corsair 460X RGB
OS: Windows 10 Ver. 21H1 Build 19043.1110


In the clip I took:

https://imgur.com/a/HmhaCyt

View: https://imgur.com/a/HmhaCyt
while measuring RPM using HWMonitor, the RPM of the card reads 60000000 RPM and can fluctuate between that a lot. What might be causing this? I removed MSI Afterburner and no help. I also cleared the Graphics drivers and did a clean install of them, with that not fixing it either. All help is appreciated!

 

[Ralston18]

Update your post to include full system hardware specs and OS information.

Include PSU: make, model, wattage, age, condition.

Are you overclocking anything? If so then provide configuration details.

 

[bigjames489]

Update your post to include full system hardware specs and OS information.

Include PSU: make, model, wattage, age, condition.

Are you overclocking anything? If so then provide configuration details.

Updated.

 

[Flayed]

6e07 RPM? Wow that card should really be flying

 

[Bazzy 505]

Don't worry, no fan spin that fast, if it did it would have been an ufo. What you have here is either bad pulse contact on one of the three fan connectors ( you can try to reseat the connectors), dry solder joint on one of those pins ( both very unlikely) or busted hall sensor on one of the fans ( most probable cause), which would cause result in described behavior. If the card is still in warranty submit a RMA ticket.

If it's out of warranty you can purchase preplacement set of fans for a few bucks here.
https://www.gpufanreplacement.com/collections/asus-graphics-card-fan-replacements/triple

 

[bigjames489]

Don't worry, no fan spin that fast, if it did it would have been an ufo. What you have here is either bad pulse contact on one of the three fan connectors ( you can try to reseat the connectors), dry solder joint on one of those pins ( both very unlikely) or busted hall sensor on one of the fans ( most probable cause), which would cause result in described behavior. If the card is still in warranty submit a RMA ticket.

If it's out of warranty you can purchase preplacement set of fans for a few bucks here.
https://www.gpufanreplacement.com/collections/asus-graphics-card-fan-replacements/triple

This sounds like whats happening. Ill try to reseat the fan connectors and as a last resort replace the fans myself. I believe Asus warranty is only a year so an RMA ticket wouldn't be possible.

 

[Colif]

its clearly wrong as its constantly moving up and down between values, nothing can spin up to 600k rpm in such a short time and then slow down a second later... that and I think you would hear it way more than a mild ramping up

 

[MrN1ce9uy]

HWMonitor is known to give false readings that are astronomically high.

Use HWInfo instead: https://www.hwinfo.com/download/

 

[Cowgoesmoo2]

You can rest assured that your reading is wrong and a result of computer error and be happy that your GPU fan is NOT running at at 60 million RPM. If you can get some kind of a fan to spin up to 60 million RPM in under a second, then you may be eligible for a nobel prize in some sort of physics that no one has heard about in human history. You have some serious technology and materials and engineering that we could use to go leave the solar system.

See if you could get a fan to spin up to 60 million RPM.. so well if we do the math, your GPU is about 230x111 millimeters, or .23x.111 meters. Looking at the GPU, in the short direction the fans are almost touching the edge of the case so let's just round that to .100, it's rather close.

Given the radius of the fans what's the velocity at the tip if it moves at 60 million RPM? The equation giving pure speed is v=wr, but w is in radians.
So convert RPM to w. That would be 6.28 million radians per second. v=wr would give 6.28 million times the radius of .1 meters, or a velocity of 628,000 meters per second.


The speed of light is 300,000,000 meters per second or 300 million meters per second, so at the very outer edge of your GPU fan, it is moving at 0.5% of the speed of light. Huh.


(I did make a mistake in assuming that the mass of winglets of the fans are assumed to all be moving at 627,000 meters per second)

Supersonic travel at sea level begins at 343 m/s which would be about 0.00054 PERCENT of the speed which the tip of your fan is spinning at.


How heavy is a wingtip of a GPU fan? They're supposed to be lightweight, so let's just say 10 or maybe 15 grams. Given the equation for kinetic energy K=0.5mv^2, (0.5)(.0015 kg)(394384000000 velocity squared)

The Kinetic Energy of ONE winglet on your GPU, if estimating with 15 grams as weight would be 295,788,000 Joules of energy. It looks like the ROG STRIX GTX 1660 Ti has nine winglets. Let's just say the center platform doesn't exist- 2,662,092,000 Joules.

Wikipedia lists

1–10×109 J Energy in an average lightning bolt[116] (thunder) 1.1×109 J Magnetic stored energy in the world's largest toroidal superconducting magnet for the ATLAS experiment at CERN, Geneva[117] 1.2×109 J Inflight 100-ton Boeing 757-200 at 300 knots (154 m/s) 1.4×109 J Theoretical minimum amount of energy required to melt a tonne of steel (380 kWh)[118][119] 2×109 J Energy of an ordinary 61 liter gasoline tank of a car.[96][120][121] 2×109 J The unit of energy in Planck units[122] 3×109 J Inflight 125-ton Boeing 767-200 flying at 373 knots (192 m/s)

(The 109s are supposed to be 10^9 but the formatting doesn't work)
So your GPU would be gaining and losing a bit as much energy as a slow-flying Boeing 767 within seconds. So your GPU fans (there are two of them, actually) could be spun up to launch and receive Boeing 767s within seconds. And we're ignoring the part where the center platform of the fan exists at all.

So yeah, if your fan spun that quickly, you would create sonic winds (probably winds that move far too fast to be called winds) inside your computer case, with some airflow or compressions possibly traveling at speeds of half a percent of the speed of light in some regions. At those speeds, I wonder if airflow would even stay in a gaseous form of matter.