Your new build powers on, the lights come up, Windows boots, but the GPU fans sit perfectly still. Or maybe your long-standing card suddenly shows the same behavior after a driver update, and every game session ends in thermal throttling or a shutdown. Either way, a graphics card with stationary fans at boot looks alarming, and many users assume the card is dead on arrival.
Here is the good news: in many cases, this is not a failure at all. Most modern GPUs use a zero-RPM mode that keeps fans off until the core reaches roughly 50 to 60 degrees Celsius. However, if your fans never spin even under load, or if you have disabled zero-RPM in software and they still stay silent, something needs fixing. Work through the fixes below in order; most readers will solve the problem within the first three. None of these steps take more than about 20 minutes, and only the last few require opening the card itself.
Likely Causes
| Cause | How to confirm | Fix difficulty |
|---|---|---|
| Zero-RPM / semi-passive mode is active by design | Fans stay off at idle but spin up once GPU temperature exceeds about 55°C under load | None — normal behavior |
| Fan curve set too low or disabled in tuning software | Check the fan tab in MSI Afterburner or the vendor app; the curve sits at 0% until high temps | Easy |
| Outdated, corrupted, or mismatched display driver | Fan control stopped working after a driver update; Device Manager shows a warning icon on the GPU | Easy |
| PCIe power cables not fully seated | Reseat the 8-pin/12V-2×6 connectors and the fans begin spinning | Easy |
| Stiction — fan bearings seized after long idle periods | Fans twitch but do not turn; a gentle flick frees them temporarily | Moderate |
| Failed fan or blown fan header on the card’s PCB | Fans receive no voltage even with a forced 100% curve; requires multimeter testing | Hard |
Fix 1: Confirm It Is Actually Zero-RPM Mode
What causes it
Nearly every GPU released in the last several years includes a semi-passive fan mode. The BIOS on the card holds the fans at 0% until the die reaches a threshold temperature, typically 50–60°C, to eliminate noise at idle and extend fan lifespan.
How to check
Open MSI Afterburner (free from MSI’s site), enable hardware monitoring, and watch the “FAN SPEED” and “GPU TEMPERATURE” readouts while running a load such as the Heaven benchmark or a demanding game. Alternatively, press Ctrl+Shift+Esc to open Task Manager, go to the Performance tab, select GPU, and watch the temperature while running any 3D application.
What to do
If the fans spin up once the temperature crosses roughly 55°C and stay spinning while the card is hot, your card is functioning exactly as designed. No fix is needed. If you want the fans to run at idle for testing peace of mind, see Fix 2.
How to undo it
Nothing to undo — you have not changed anything. If you later forced a fan curve, simply click the reset icon (circular arrow) in Afterburner to return the card to its stock vBIOS fan profile.
Fix 2: Override the Fan Curve in Afterburner
What causes it
A user profile, vendor app preset (such as ASUS GPU Tweak’s “Silent” or “Quiet” profile), or a default curve that holds 0% until a high temperature can make fans appear dead when they are simply being told not to spin.
How to check
Open MSI Afterburner, click the gear icon to open Properties, select the Fan tab, tick “Enable user defined software automatic fan control,” and inspect the graph. If the curve is pinned at 0% across the low-temperature range, the software is holding the fans off. Also check vendor utilities: in ASUS GPU Tweak, look under the Fan tab; in Gigabyte Control Center, check the VGA fan section; in NVIDIA App or AMD Adrenalin, review any custom performance profiles.
What to do
In Afterburner’s fan curve editor, drag the control points so the curve reads approximately 30% at 40°C, 50% at 60°C, and 80% at 75°C. Click Apply (the checkmark button), then confirm the fans spin at idle or under mild load. Only one fan-control application should be running at a time — close vendor utilities while testing with Afterburner, since two apps fighting over fan control can force 0%.
How to undo it
Untick “Enable user defined software automatic fan control” in the same Fan tab and click Apply. The card reverts to its factory curve and zero-RPM behavior.
Fix 3: Reseat the PCIe Power Connectors
What causes it
An 8-pin PCIe or 12V-2×6 (12VHPWR) connector that is not fully clicked in can deliver enough power for basic boot but leave the card operating in a degraded state where fan control misbehaves or the card limits itself.
How to check
Shut down, flip the PSU switch off, and press the case power button for five seconds to drain residual charge. Unplug and firmly reconnect each power cable at the GPU end until the latch clicks. For 12V-2×6 connectors, verify the cable is inserted square and flush — a partially seated 16-pin connector is a known fire and malfunction risk, so inspect the seating from the side.
What to do
Avoid daisy-chained pigtails on high-wattage cards; use two separate PCIe cables from the PSU for cards drawing over 250 W. Power back on and test with a game or benchmark while watching fan RPM in Afterburner.
How to undo it
No changes persist — reseating cables is reversible by nature. If you added a second cable, you can return to the pigtail, but keeping the separate cables is strongly recommended.
Fix 4: Clean-Reinstall the Graphics Driver
What causes it
A corrupted or partially applied driver can disrupt communication between Windows and the GPU’s fan controller, particularly after an in-place upgrade or a failed driver install.
How to check
Right-click Start, choose Device Manager, expand Display adapters, and look for a yellow warning triangle on your GPU. Also note whether fan control suddenly broke after a recent driver update — that timing points to a driver problem.
What to do
Download Display Driver Uninstaller (DDU), boot into Safe Mode (hold Shift while clicking Restart, then navigate Troubleshoot > Advanced options > Startup Settings > Restart, and press 4), and run DDU selecting “Clean and restart” for your GPU vendor. After the reboot, install the latest driver directly from NVIDIA or AMD. During installation, choose “Custom” and tick “Perform a clean installation.”
How to undo it
Driver clean installs do not need undoing. If the new driver misbehaves, use Device Manager > Display adapters > your GPU > Properties > Driver tab > Roll Back Driver, or reinstall a previous driver version downloaded from the vendor’s archive.
Fix 5: Update the GPU’s vBIOS
What causes it
Some cards shipped with early vBIOS versions that had fan-control bugs — fans failing to spin up at expected temperatures, or zero-RPM thresholds set incorrectly. Board partners have issued corrected vBIOS updates for several models.
How to check
Open GPU-Z and read the “BIOS version” field on the Graphics Card tab. Visit your card manufacturer’s support page, search your exact model, and compare against the latest listed vBIOS. Check the release notes for mentions of fan behavior.
What to do
Download the vendor’s vBIOS update utility (ASUS GPU Tweak includes a vBIOS flasher; other brands ship standalone flash tools). Run the tool as administrator, follow the on-screen prompts exactly, and do not power off or reboot during the flash — interrupting it can brick the card. Ideally, flash from a system with a second display source or accept that the screen may blank briefly.
How to undo it
Most flash tools support reflashing the original BIOS if you saved a backup copy when prompted. Some dual-BIOS cards (common on high-end models) have a physical switch near the PCIe power connectors; flipping it loads the backup BIOS instantly.
Fix 6: Free Seized Fan Bearings (Stiction)
What causes it
A card stored for months, or a secondary card used infrequently, can develop stiction — the sleeve or fluid-dynamic bearing lubricant settles and the motor lacks enough starting torque to break the rotor loose.
How to check
With the system off and the PSU switched off, look at the fans through the case side panel or shroud. If they twitch slightly when the system powers on but never rotate, that is classic stiction. A fan that does not move at all may be electrically dead instead (see Fix 7).
What to do
Remove the card and lay it on an anti-static surface. Gently spin each fan blade with a fingertip — it should rotate freely with slight resistance. If it feels gritty or locked, apply a single small drop of machine oil at the bearing hub (the center cap under any sticker, if accessible), then rotate the blade by hand for 30 seconds to distribute the lubricant. Reinstall and test with a forced 100% fan curve in Afterburner. Never lubricate while the card is powered.
How to undo it
Blot away any excess oil with a lint-free cloth and reseat the card. No software state changed, so nothing else to revert.
Fix 7: Test Whether the Fan or the Card’s Fan Header Has Failed
What causes it
The fan motor, its cable, or the PWM driver circuit on the card’s PCB can fail outright, often after a surge, liquid spill, or years of dust accumulation.
How to check
In Afterburner, force the fan to 100% and listen at the card with the side panel off. If you hear no motor sound at all and the blades do not twitch, the fan is getting no drive. If you have a multimeter, with the system off, disconnect the fan’s small connector from the PCB and check for continuity across the fan’s power pins on the fan side — an open circuit indicates a dead fan. Testing voltage at the card’s header requires the system running and care around live components.
What to do
If the fan is dead and the card is under warranty (typically 2–3 years for board partners like EVGA, ASUS, MSI, and Sapphire), start an RMA — do not open the shroud, as stickers breaking void coverage. Out of warranty, replacement fans specific to your card model are sold by parts resellers, or you can attach a high-CFM case fan blowing directly over the card as a stopgap while monitoring temperatures closely; abort any session if the GPU exceeds 85°C.
How to undo it
Disconnect any stopgap fan and restore the card’s original fan connector if you replaced hardware. Reinstall the stock shroud and screws in their original positions.
If Nothing Works
If you have confirmed a working driver, a forced 100% fan command, and seated power connectors, and the fans still receive no power, the fault is almost certainly on the card’s PCB — typically the fan controller IC or a failed header circuit. At that point, an RMA is the correct path if warranty remains; otherwise, a repair shop with board-level capability can often replace the fan controller for far less than a new card. As an interim measure, cards can run passively for light desktop work, but gaming without functioning fans will hit thermal limits within minutes. As a last diagnostic, test the card in a second PC — if the fans work there, your primary system’s PCIe slot, riser cable, or PSU rails are the real culprit, and swapping those components is your fix.
FAQ
Is it bad if my GPU fans never spin at idle?
No. Zero-RPM mode is intentional on virtually all modern cards. Fans engaging only under load is the designed behavior and actually extends fan lifespan. Only worry if they never spin even when temperatures climb past 60°C.
What temperature should GPU fans start spinning at?
Factory zero-RPM thresholds usually sit between 50 and 60°C. Below that, the card relies on its heatsink alone. You can override this with a custom fan curve in MSI Afterburner if you prefer active cooling at all times.
Can I force my GPU fans to spin constantly?
Yes. In MSI Afterburner, open Properties > Fan tab, enable user-defined fan control, and set the curve to a flat 30–40% from 30°C upward. Expect slightly higher noise and marginally more dust accumulation, but no harm to the card.
Will a GPU overheat if the fans don’t spin during startup?
Not during startup — a cold card at the desktop draws little power and idle temperatures of 30–45°C are safe passively. The danger only exists under sustained load, where a card without functioning fans will exceed 90°C and throttle or shut down within minutes.
