DRIVER_POWER_STATE_FAILURE (stop code 0x9F) means a device driver failed to complete a power transition—usually during sleep, wake, hibernate or Fast Startup—within Windows’ timeout. Check Parameter 1 in Event Viewer (Kernel-Power 41): 0x3 means a device blocked a power IRP, while 0x4 means a power transition timed out waiting on the Plug-and-Play lock. Turn off Fast Startup, set PCIe Link State Power Management to Off as a test, and update or roll back your Wi-Fi, GPU, chipset and NVMe drivers.
At a glance
- Applies to
- Windows 11 and Windows 10, stop code 0x0000009F
- Typical trigger
- Sleep, wake, hibernate, shutdown, or Hybrid Shutdown (Fast Startup)
- First check
- BugcheckParameter1 in Event Viewer (0x3 blocked IRP vs 0x4 PnP timeout)
- Fastest test
- Disable Fast Startup and PCIe Link State Power Management
Provenance
- Nov 21, 2026Published
Despite the name, DRIVER_POWER_STATE_FAILURE (stop code 0x0000009F) is almost never a bad power supply. Do not order a replacement PSU for this blue screen.
Microsoft’s 0x9F bug-check reference states what actually failed: a device driver entered an inconsistent or invalid power state—almost always because Windows told a piece of hardware to go to sleep, wake up, hibernate, or power down, and the driver holding that device failed to answer the power request before Windows’ watchdog timer expired. That is why 0x9F characteristically strikes two to five minutes after you put a PC to sleep, close a laptop lid, wake from idle, or shut down.
Step 1. Read Parameter 1 in Event Viewer (no download needed)
Before changing settings at random, check which 0x9F subcode Windows logged.
- Right-click Start, open Event Viewer, and navigate to Windows Logs > System.
- Click Filter Current Log… on the right, enter Event IDs
41, 1001, and click OK. - Open the Kernel-Power (Event ID 41) entry from the crash timestamp and click the Details tab (
Friendly VieworXML View):BugcheckCode159in decimal equals0x9Fin hex (0means sudden power loss—for that case, switch to PC randomly restarts).- Read
BugcheckParameter1.
Match Parameter 1 against Microsoft’s official 0x9F table:
Parameter 1 (Arg1) | Official Microsoft violation | What hung & what Arg2/Arg4 hold |
|---|---|---|
0x3 (most common) | A device object has been blocking a power IRP (I/O Request Packet) for too long | A PCIe, USB, Wi-Fi, GPU or NVMe driver refused to sleep or wake. Arg2 is the device’s Physical Device Object (PDO); Arg4 is the blocked IRP |
0x4 | Power state transition timed out waiting to synchronize with the Plug-and-Play (PnP) subsystem | A driver thread held the PnP lock across a sleep/wake or hot-plug event (Arg2 = timeout in seconds; Arg3 = stalling thread) |
0x5 / 0x6 | Directed power transition (PoFx Modern Standby) timed out (0x5) or failed its callback (0x6) | Modern Standby (S0 Low Power Idle) device driver bug on laptops and newer desktops |
0x1 / 0x2 / 0x500 | Device object freed with an uncompleted power request (0x1), or completed a system power IRP without calling PoStartNextPowerIrp (0x2 / 0x500) | Legacy filter driver, USB dongle, or virtual adapter driver bug |
Look at the System log entries 30 to 180 seconds before the Kernel-Power 41 event too. An error from Netwtw14 (Intel Wi-Fi), mtkwl6ex (MediaTek Wi-Fi), nvlddmkm (NVIDIA), amduw23g (AMD), stornvme (NVMe SSD), or USBXHCI right before the reboot often names the offending device without opening a debugger.
Step 2. Stop the sleep/shutdown loop: disable Fast Startup & PCIe Link State Power Management
If 0x9F hits every time you shut down, cold-boot, or leave the desk for ten minutes, stabilize the machine first with two built-in Windows power toggles.
1. Turn off Windows Fast Startup (fixes shutdown & boot 0x9F)
Windows “Shut down” defaults to a hybrid hibernate (S4) session for kernel drivers, carrying hung driver power states across reboots.
- Press
Win + R, typepowercfg.cpl, and press Enter. - Click Choose what the power buttons do on the left.
- Click Change settings that are currently unavailable.
- Uncheck Turn on fast startup (recommended) and click Save changes.
- Run one real full restart (Start > Power > Restart) to flush the saved kernel state.
2. Turn PCIe Link State Power Management (ASPM) off as a test (fixes idle & sleep 0x9F)
Aggressive PCIe Active State Power Management (ASPM) puts discrete GPUs, Wi-Fi cards, and secondary NVMe drives into deep L1 link sleep states that buggy firmware or drivers fail to wake from, triggering Parameter 1 = 0x3.
- In
powercfg.cpl, click Change plan settings next to your active power plan, then click Change advanced power settings. - Expand PCI Express > Link State Power Management.
- Set Setting (or both On battery and Plugged in on a laptop) to Off, then click Apply.
- Use the PC normally across two or three sleep/wake cycles. If
0x9Fdisappears completely, you just confirmed a PCIe device sleep handshake bug.
Step 3. Pinpoint the exact .sys driver in 3 minutes with WinDbg
Unlike memory corruption crashes where random drivers take the blame, 0x9F with Parameter 1 = 0x3 or 0x5 records the exact hardware device stack (PDO) that refused to answer. Reading it takes three minutes and removes all guesswork:
- Install WinDbg from the official Microsoft Store (free, published by Microsoft Corporation).
- Copy the newest
.dmpfile fromC:\Windows\Minidumpto your Desktop (so WinDbg doesn’t need elevated folder permissions) and open it in WinDbg. - At the bottom command bar, type
!analyze -vand press Enter. - Read two lines in the output:
IMAGE_NAME/MODULE_NAME(ordx KiBugCheckDriver): if it names a third-party driver such asrtwlane6.sys,nvlddmkm.sys, origdkmdn64.sys, that driver stalled the power request.- If
IMAGE_NAMEonly sayspci.sys,acpi.sys, orntkrnlmp.exe, Windows is naming the bus driver sitting underneath the hung hardware. Copy the hex address fromArg2(Physical Device Object of the stack) and run:!devstack <Arg2-address> - Look at the
DriverNameandInstancePathcolumn (for example,PCI\VEN_10DE...for an NVIDIA GPU,PCI\VEN_8086...for an Intel Wi-Fi/Thunderbolt controller, orPCI\VEN_144D...for a Samsung NVMe drive). Match that Vendor/Device ID in Device Manager > Properties > Details > Hardware Ids to see the exact component name in plain English.
Step 4. Fix the four most common 0x9F device culprits
Once Step 1 or Step 3 names the subsystem, fix that driver directly:
- Discrete GPU (
nvlddmkm.sys,amdkmdag.sys,dxgkrnl.sys— common on gaming laptops switching between iGPU and dGPU):- Clean-install the current WHQL graphics driver from NVIDIA, AMD, or your laptop OEM’s support page—follow our GPU driver crash & clean-install guide so leftover driver store state doesn’t persist.
- Wi-Fi / Bluetooth / Ethernet adapter (
Netwtw*.sys,mtkwl6ex.sys,rtwlane*.sys,e2fnexpress.sys):- Download the latest WLAN + Bluetooth + LAN drivers from your motherboard or laptop maker’s support page.
- In Device Manager > Network adapters, double-click your Wi-Fi or Ethernet card, open the Power Management tab (if exposed outside Modern Standby), and uncheck Allow the computer to turn off this device to save power.
- USB hubs, Thunderbolt docks, Audio interfaces, or RGB controllers (
USBXHCI.SYS,ucx01000.sys,UsbHub3.sys):- Unplug external USB hubs, KVMs, capture cards, and docks overnight as an isolation test. If
0x9Fstops while unplugged, update the dock’s firmware and your motherboard’s AMD Chipset Drivers (fromamd.com) or Intel Chipset / Thunderbolt drivers.
- Unplug external USB hubs, KVMs, capture cards, and docks overnight as an isolation test. If
- NVMe SSD (
stornvme.sys,iaStorVD.sys):- Check the drive’s firmware version in the manufacturer’s SSD toolbox (Samsung Magician, SanDisk/WD Dashboard, Crucial Storage Executive, Kingston SSD Manager) and run the quick health check in our SSD health check guide. Several PCIe 4.0 and 5.0 NVMe drives shipped firmware updates specifically addressing
D3sleep-state wake timeouts.
- Check the drive’s firmware version in the manufacturer’s SSD toolbox (Samsung Magician, SanDisk/WD Dashboard, Crucial Storage Executive, Kingston SSD Manager) and run the quick health check in our SSD health check guide. Several PCIe 4.0 and 5.0 NVMe drives shipped firmware updates specifically addressing
Step 5. Update BIOS/AGESA and repair system files
ACPI sleep tables (S3 sleep and S0ix Modern Standby) come from your motherboard UEFI firmware. If chipset and device drivers are current and 0x9F still triggers on wake:
- Check your motherboard or laptop support page for a newer BIOS / UEFI revision mentioning ACPI, S3/Modern Standby, or PCIe power management—use our step-by-step BIOS update walkthrough.
- Because repeated hard hangs during shutdown or sleep can leave Windows system files damaged, open Terminal (Admin) and run Microsoft’s standard DISM + System File Checker repair pair once the crashes stop:
DISM /Online /Cleanup-Image /RestoreHealth sfc /scannow
If your machine throws different stop codes on every crash (0x1A, 0x50, 0x139) rather than 0x9F consistently around sleep or shutdown, step back to the Diagnostic Workbench or test memory stability in Fix MEMORY_MANAGEMENT (0x1A).
Frequently asked questions
What causes DRIVER_POWER_STATE_FAILURE (0x9F)?
Does DRIVER_POWER_STATE_FAILURE mean my power supply (PSU) is dying?
Why does 0x9F happen when I close my laptop lid or put my PC to sleep?
How do I find out which exact driver caused 0x9F?
Does turning off Fast Startup fix DRIVER_POWER_STATE_FAILURE?
Sources and corrections
- Last updated
- Methodology
- See our methodology for research and review standards. It draws on 4 cited sources, listed below, each checked against the original page on the date above. All stop-code definitions, Parameter 1 through Parameter 4 meanings, and WinDbg triage structures (nt!_TRIAGE_9F_POWER, nt!_TRIAGE_9F_PNP, KiBugCheckDriver) come directly from Microsoft Learn’s Bug Check 0x9F reference opened November 21, 2026 (UTC), paired with Microsoft’s Event ID 41 and stop-code troubleshooting documentation.
- Sources
- Microsoft Learn: Bug Check 0x9F DRIVER_POWER_STATE_FAILURE accessed 2026-11-21
- Microsoft Learn: Event ID 41 when the system restarts unexpectedly accessed 2026-11-21
- Microsoft Learn: Analyze bug check (stop code error) data accessed 2026-11-21
- Microsoft Support: Troubleshooting Windows unexpected restarts and stop code errors accessed 2026-11-21
- Corrections
- Spotted an error? Email a correction with the affected line and what you found. The draft includes this page’s address. Verified changes are recorded here. Contact details.