Your AIO was running fine, then you flashed a BIOS update, rebooted, and now the radiator fans are dead silent. The pump might be humming along, the CPU might be at idle temperatures, but the fans on the radiator refuse to spin — or they twitch once at boot and stop. Sometimes only one fan header is affected, sometimes everything plugged into the CPU_FAN and SYS_FAN headers goes quiet at the same time. The screen still works, Windows still loads, so it’s tempting to assume the fans just died at an inconvenient moment.
Don’t ignore it. A BIOS update is almost never a coincidence here — it typically resets every fan curve, re-enables Q-Fan or Smart Fan control defaults, or flips a header from PWM to DC mode (or vice versa), leaving your fans with a PWM signal they can’t use or a duty cycle pinned at zero. With the radiator fans offline, your coolant loop has nowhere to dump heat. Under light desktop use you might idle at 40–55°C and never notice, but the moment you launch a game or run a render, CPU temperatures can spike to throttle points or thermal shutdown within seconds to a couple of minutes. Treat this as urgent: fix it before doing any heavy workloads, and keep an eye on CPU temperature with HWiNFO64 in the meantime.
Likely Causes
| Cause | How to confirm | Fix difficulty |
|---|---|---|
| BIOS update reset fan curves to silent/zero-RPM defaults | Enter BIOS, open the fan control page (Q-Fan Control, Smart Fan, or Fan Control depending on board) and check if CPU/radiator headers show 0% duty at low temps | Easy |
| Header mode switched between PWM and DC | BIOS fan page shows header mode as “DC” for a 4-pin PWM fan, or “PWM” for a 3-pin DC fan | Easy |
| Fans are plugged into a header the BIOS now treats as disabled or reassigned (e.g., pump header takeover) | Check whether the AIO_FAN/PUMP header settings changed; verify which physical headers your fans actually connect to | Easy |
| Zero-RPM / Fan Stop feature enabled by default after update | Fans spin briefly at power-on, then stop; BIOS shows “Fan Stop” or 0 RPM minimum enabled below ~40–60°C | Easy |
| Fan curves now reference the wrong temperature source | Curve source is set to a motherboard/PCH sensor instead of “CPU” so duty stays at minimum | Easy |
| Firmware-level header fault or corrupted update | Fans dead in BIOS on every header, including known-good fans moved between headers | Hard |
Fix 1: Reconfigure Fan Curves in BIOS
What causes it
A BIOS update wipes NVRAM-stored fan settings and restores defaults, which are frequently tuned for silence: a low minimum duty cycle (20–30%) or a flat curve that never ramps until temperatures are already high. If your radiator fans were previously on an aggressive curve, they may now be spinning so slowly (or stopped) that they’re easy to miss.
How to check
Reboot and press Del (most boards), F2 (many laptops and some ASRock/Dell systems), or F1 (Lenovo) to enter BIOS. Open the fan page — “Q-Fan Control” on ASUS, “Smart Fan” or “Hardware Monitor” on MSI and Gigabyte, “FAN-Tastic Tuning” on ASRock. Look at each radiator fan header: note the current RPM, the duty percentage, and the shape of the curve.
What to do
Select each fan header attached to the radiator and set a manual curve with the temperature source set to CPU (or coolant/liquid temp if your AIO reports it). Set the minimum duty to at least 30–40% so fans always spin audibly at idle, and ramp to 100% by 75–80°C. On ASUS boards, disable “Fan Stop” for these headers; on MSI, set fan stop temperature off or above 70°C. Save with F10 and confirm fans spin at the BIOS hardware monitor page before booting Windows.
How to undo it
Return to the fan page and choose “Restore Defaults” or re-run the board’s auto-tuning utility (ASUS Fan Xpert, MSI Fan Tune) if you preferred the previous behavior.
Fix 2: Set the Correct PWM/DC Header Mode
What causes it
After a reset, headers may revert to “Auto” detection and misidentify your fans. A PWM fan on a DC-mode header gets no usable drive voltage; a 3-pin DC fan on a PWM-mode header may receive no duty signal and sit still. Both look identical from the outside: dead fans.
How to check
In the BIOS fan page, find the mode setting per header — usually labeled “PWM,” “DC,” or “Auto.” Count the wires on your fan connectors: 4 wires (PWM), 3 wires (DC). Check the mode shown for CPU_FAN, CPU_OPT/CHA_FAN, and any AIO_PUMP or AIO_FAN header your fans use.
What to do
Set each radiator fan header explicitly to PWM if the fans are 4-pin, or DC if 3-pin. Avoid “Auto.” If your fans plug into the AIO’s own PWM pass-through cable, that cable ultimately feeds a motherboard header, so fix the mode there. Save and reboot; fans should now respond to the curve from Fix 1.
How to undo it
Switch the header back to Auto or its previous mode in the same menu.
Fix 3: Disable Zero-RPM / Fan Stop
What causes it
Many boards and AIO controllers default to a zero-RPM mode after a settings reset: fans stay completely stopped until the temperature source crosses a threshold, often 50–60°C. Because radiator fans often track coolant temperature (which rises slowly), they can appear dead for a very long time even under load.
How to check
In BIOS fan settings, look for options named “Fan Stop,” “Zero RPM,” “Fan Off,” or a stop-temperature field set around 40–60°C. In Windows, watch HWiNFO64: if fans read 0 RPM while CPU temperature climbs past 60°C under load, fan stop is active somewhere.
What to do
Disable Fan Stop for every radiator fan header, or raise the stop threshold above your maximum expected coolant temperature (80°C is a safe ceiling for the setting; most coolant stays under 60°C). Set a nonzero minimum duty (30%+) as in Fix 1. If your AIO has its own software (e.g., a pump/fan controller app from the cooler vendor), open it and disable any silence/zero-RPM profile there too — the BIOS setting doesn’t override an in-firmware cooler profile.
How to undo it
Re-enable Fan Stop and set your preferred stop temperature in the same BIOS page or cooler software.
Fix 4: Clear CMOS Properly After the Update
What causes it
A partially applied or interrupted settings migration during the update can leave the Embedded Controller and fan controllers in a bad state — RPM sensors reading 0, headers stuck off. A full CMOS reset forces the firmware to re-enumerate all hardware monitoring.
How to check
This applies if Fixes 1–3 didn’t help and fans are dead even in BIOS on every header, or RPM readings show “N/A”/0 across the board.
What to do
Shut down, flip off the PSU switch, and unplug the AC cord. Press the power button for 10 seconds to drain residual power. Either remove the coin-cell battery for 5 minutes or short the CLR_CMOS/CLRTC jumper/pads per your manual (many boards have a dedicated “Clear CMOS” button on the rear I/O — hold it 5–10 seconds with the system powered off). Restore power, boot into BIOS, and reconfigure fan modes and curves from Fixes 1–3. Note this also resets boot order and XMP/EXPO, so re-enable your memory profile afterward.
How to undo it
Not applicable — CMOS reset simply restores defaults; reapply your preferred settings.
Fix 5: Replug Fans Onto a Known-Good Header
What causes it
Sometimes the specific header fails to initialize after firmware changes, or fans were daisy-chained through the AIO’s pump body or a controller that lost its configuration. Moving the connection isolates the fault.
How to check
Unplug one radiator fan and connect it directly to CPU_FAN (or CPU_FAN1). Power on and enter BIOS. If it spins there but not on its original header (e.g., SYS_FAN4 or a splitter), the problem is header- or chain-specific. Also test each fan individually to rule out one bad fan dragging down a daisy chain.
What to do
If fans work on CPU_FAN, either leave them there or move them to a different CHA_FAN header that tests healthy. If you use a splitter or hub, connect it to an AM4/AM5/LGA1700 board header rated for the total amperage — most motherboard headers supply 1–2 A; check your manual. Reapply PWM/DC mode and curve settings for the new header.
How to undo it
Move fans back to the original header and confirm it now initializes after the earlier fixes.
Fix 6: Reflash or Roll Back the BIOS
What causes it
A corrupted or buggy firmware release can break Super I/O fan-channel initialization entirely. This is rare but real; if fans are dead on all headers with known-good fans and settings, the firmware itself is suspect.
How to check
Confirm you’ve exhausted Fixes 1–5. Test multiple fans on multiple headers with zero response, and fan RPM fields dead in BIOS. Note your current BIOS version (shown on the BIOS main page or with winver-style tools like msinfo32 under “BIOS Version/Date”) and check whether others on your board’s model report the same issue on that version.
What to do
Download the same BIOS version again from your board vendor’s support page and reflash it via the BIOS update utility (EZ Flash on ASUS via F7 → EZ Flash; M-Flash on MSI; Q-Flash on Gigabyte) from a FAT32 USB stick, even if it’s already installed — this rewrites the firmware image. Alternatively, install the immediately previous version to roll back, if your board permits downgrades. Do not power off during the flash; use a UPS or wait for stable grid conditions. Afterward, redo Fixes 1–3 since settings reset again.
How to undo it
Flash a different version (forward or back) using the same procedure.
Fix 7: Update or Clean-Reinstall the Vendor Software Layer
What causes it
On some platforms, fan behavior is partly managed by vendor utilities and drivers (ASUS Armoury Crate, MSI Center, Gigabyte Control Center) that install kernel-level services. A BIOS update can desynchronize these from the new firmware, leaving software profiles overriding or zeroing fan output.
How to check
If fans behave in BIOS but stop once Windows loads, software is the culprit. Boot into BIOS and let it sit at the hardware monitor page: fans spinning there but dead in Windows points squarely at OS-level interference.
What to do
Open the vendor suite (e.g., Armoury Crate → Fan Xpert, or MSI Center → User Scenario / Hardware Monitor), check for an app update, and reapply a fan profile. If behavior is still wrong, uninstall the suite fully using the vendor’s removal tool (ASUS offers a dedicated Armoury Crate Uninstall Tool; MSI Center has an uninstall component in its installer), reboot, and either reinstall fresh or run bare without the software and rely on BIOS curves only.
How to undo it
Reinstall the utility and restore your saved profile.
If Nothing Works
If fans are unresponsive on every header, with multiple known-good fans, after a full CMOS reset and firmware reflash, you’re likely dealing with a hardware failure — a dead fan header, blown fan controller on the Super I/O chip, or a pump/fan hub that died. At that point, contact your motherboard manufacturer’s support with your board model, BIOS version, and the troubleshooting steps you’ve taken; header failure under warranty is a legitimate RMA reason. As a stopgap, a powered fan hub connected to USB and SATA power can run your radiator fans independently of the motherboard headers, but monitor CPU temperatures closely with HWiNFO64 and set a 90°C alarm until the underlying fault is resolved. If temperatures climb past 85–90°C at idle, shut down immediately.
FAQ
Why would a BIOS update stop my fans at all?
Fan control lives in firmware. An update resets NVRAM settings to factory defaults and reinitializes the fan controller, which commonly wipes custom curves, changes PWM/DC modes, or enables zero-RPM silent behavior — any of which can leave radiator fans stopped even though nothing is physically broken.
My pump is spinning but the radiator fans aren’t. Is the AIO dead?
Almost certainly not. The pump and fans are controlled separately — the pump usually sits on a dedicated header running at fixed full speed, while fans follow a curve. A dead pump means rapid, aggressive overheating within seconds of load; fans-only failure points to the header and curve issues described in this guide.
Is it safe to run the CPU while the radiator fans are off?
Only briefly. At idle, coolant heats slowly and you may be fine for several minutes. Under load, an AIO with stagnant coolant can push a CPU to 90–100°C quickly, triggering throttling or shutdown. Watch temperatures in HWiNFO64, avoid heavy workloads, and shut down above 90°C.
Should I update the BIOS again or wait for a newer version?
First reflash the same version — a clean reflash fixes corrupted images without introducing new behavior. If the fault persists on that specific release, roll back to the previous version if your board allows it, and only move forward again once the vendor ships a newer release, checking the changelog for hardware monitor or fan control fixes.
