STATUS_GRAPHICS_WRONG_ALLOCATION_DEVICE (0xC01E0115) Fix

Had a client last month — video editor, dual-GPU Lenovo P1 — whose machine started throwing 0xC01E0115 every time she scrubbed a 4K timeline in Premiere. Not on boot. Not when idle. Only under load, and only after the Windows Update that pushed a newer Intel Iris driver alongside her existing NVIDIA Studio driver. Classic pattern.

STATUS_GRAPHICS_WRONG_ALLOCATION_DEVICE (0xC01E0115) is what the Dxgkrnl.sys kernel component reports when a graphics allocation gets handed to a device that doesn't own it. In plain talk: the driver stack got confused about which GPU owns a chunk of video memory. On single-GPU desktops it happens too, usually after a botched driver install or a corrupted DirectX shader cache.

Work down this list. Stop when the BSOD or app crash quits happening.

30-Second Fix: Force a Clean GPU Reset

Before you touch drivers, do the cheap thing. Windows keeps stale GPU context alive long after a driver update, and that's often the whole problem.

1. Save everything. You're about to force a display reset.
2. Hit Win + Ctrl + Shift + B. Screen flashes black for a second, audio beeps, and the graphics stack restarts.
3. Reboot the machine normally — don't skip this part, the hotkey alone isn't enough.
4. Re-launch whatever triggered the error. Game, CAD app, video editor — whatever caused it last time.

If it crashes again immediately, this isn't a context problem. Keep going.

5-Minute Fix: Kill the Shader Cache and Reinstall the Right Driver

Nine times out of ten with 0xC01E0115 on a hybrid-graphics laptop, someone installed two drivers that disagree about which GPU handles which task. The fix is a full clean install of both, in the right order.

Step 1: Nuke the shader cache

Corrupted shader cache after a driver swap is a documented trigger for this error. Delete it:

Press Win + R, paste this, hit Enter:

%localappdata%\D3DSCache

Select everything inside and delete it.

Then also run Disk Cleanup (cleanmgr.exe) and check
"DirectX Shader Cache" before cleaning.

Step 2: Clean-install both GPU drivers

On laptop with Intel + NVIDIA: you need both drivers, and NVIDIA's installer has a checkbox you can't skip. Grab the current Intel DCH driver from Intel's site (not Windows Update), and the current Game Ready or Studio driver from NVIDIA. Then:

1. Disconnect from the internet. Seriously. Windows Update will try to shove a "recommended" driver in mid-install.
2. Run DDU (Display Driver Uninstaller) in Safe Mode. Select GPU → NVIDIA, clean and restart. Then run it again for Intel.
3. Boot back in, install the Intel driver first, reboot.
4. Install the NVIDIA driver. In the installer choose Custom, check Perform a clean installation. Reboot.
5. Reconnect the internet after both installs are done.

On single-GPU desktops you can skip the Intel step. Just DDU the one GPU driver and reinstall fresh.

Step 3: Set the preferred GPU for the crashing app

Windows' Graphics Settings panel is where hybrid-graphics apps get told which GPU to use. If it's wrong, you get this error under load.

Settings → System → Display → Graphics

Find your app in the list (or Browse to add the .exe)
Click it → Options → set to "High performance"

For laptops, also check in your vendor's control panel:
- NVIDIA Control Panel → Manage 3D Settings → Program Settings
- AMD Adrenalin → Gaming → Graphics → select the .exe

Reboot after this. Then test.

15-Minute Fix: Registry Check, DX Repair, and Driver Rollback

If the clean install didn't take, something deeper is going on. Three things left to try.

Check for a stuck driver on the wrong device

I've seen 0xC01E0115 happen because Windows still has a phantom display adapter entry pointing at the wrong GPU. Fix it in Device Manager:

1. Win + X → Device Manager.
2. View menu → Show hidden devices.
3. Expand Display adapters. Any greyed-out entries? Right-click → Uninstall.
4. Also expand Monitors and remove greyed-out phantom monitors.
5. Reboot. Windows will re-enumerate the real hardware.

Repair DirectX and the runtime

The allocation calls come through DXGI and D3D, so a broken runtime will produce exactly this error code. Run these from an admin command prompt:

DISM /Online /Cleanup-Image /RestoreHealth
sfc /scannow

Then install the DirectX End-User Runtime from Microsoft's site. Yes, even on Windows 11 — the legacy runtime DLLs are still used by older apps and their absence causes odd DXGI failures.

Roll back the driver if the crash started after an update

If the error started the same day as a Windows Update or vendor driver push, roll back instead of reinstalling forward. You want the version that worked:

1. Device Manager → Display adapters → right-click GPU → Properties.
2. Driver tab → Roll Back Driver (greyed out if no previous driver exists).
3. Reboot. Pin that driver version — use wushowhide.diagcab or Windows Update's "Pause updates" to stop Windows from re-pushing it.

Check for overclock or undervolt instability

This is the one people miss. MSI Afterburner, Intel XTU, or a vendor's "performance" profile that undervolts the GPU can cause allocation failures under load. Reset everything to default, disable XMP on the memory if it's enabled, and test. If the error goes away, your stable clock wasn't stable.

When to stop and call it hardware

If you've done the DDU clean install, rolled the driver back to a known-good version, cleared shader cache, disabled the OC, and you still get 0xC01E0115 under load — that's a bad GPU or a failing VRAM module. On a discrete card, VRAM failure often presents exactly like this because the allocation table gets corrupted when a specific memory cell misbehaves. RMA it.

One quick sanity test before you ship the card off: run Unigine Superposition in a loop for 20 minutes in a window. If it crashes with 0xC01E0115 across multiple driver versions, it's not software.

What doesn't help

• Running the DirectX Diagnostic Tool's fix suggestions. Dxdiag doesn't fix anything, it just reports. Skip it.
• Disabling hardware acceleration in Chrome. Unrelated error, you're fixing the wrong thing.
• Reinstalling Windows. Almost never needed for this code. It's a driver-layer problem, not an OS corruption problem.
• Swapping RAM. Unless memtest errors, RAM isn't your issue here.

Start at the top. Most people fix 0xC01E0115 with the DDU clean install and a reboot. The ones who don't usually find it's a bad card or a stale registry entry — both fixable, neither scary.