Skip to content

How-ToGPU

How to Turn Off CUDA System Memory Fallback for Stable Diffusion and Local AI

NVIDIA's driver can let CUDA apps spill into slower shared system memory when they run out of VRAM, which can make Stable Diffusion and local AI suddenly slow. On driver 546.01 or later, add the app's python.exe in NVIDIA Control Panel > Manage 3D settings > Program Settings and set CUDA - Sysmem Fallback Policy to Prefer No Sysmem Fallback.

By , Editorial team

3 min read

How to Turn Off CUDA System Memory Fallback for Stable Diffusion and Local AI

If image generation or a local AI app suddenly slows to a crawl on your NVIDIA graphics card, the cause may be a driver feature called system memory fallback. Instead of crashing when the GPU runs out of video memory (VRAM), the driver lets the app use shared system memory. That keeps things running, but much more slowly.

NVIDIA lets you turn this off for individual programs with the CUDA - Sysmem Fallback Policy setting. Forge PC hasn't benchmarked the setting for this guide; the steps below follow NVIDIA's own support article.

What system memory fallback does

NVIDIA explains that in driver 536.40 it added a method that allows an application to use shared memory when it has exhausted GPU memory. Apps that previously crashed when they ran out of VRAM can carry on, "albeit at lower speeds". The switch happens when the app gets close to maxing out GPU memory, so it can be seamless and you may not notice until performance drops.

NVIDIA gives Stable Diffusion as an example: it often needs close to 6GB of GPU memory, so on 6GB cards the fallback can kick in and slow generation down. In driver 546.01 and later, NVIDIA added a setting to disable the fallback. The trade-off, in NVIDIA's words, is that performance should be stable "at the risk of a crash" if you use a setting that needs more GPU memory than you have.

Should you turn it off?

  • Turn it off if you'd rather get an out-of-memory error than an unexpectedly slow run, so you know to lower the resolution, batch size or model size.

  • Leave it on if your workflows only occasionally go over your VRAM and you'd rather finish slowly than crash.

To see whether you're going over your VRAM, our guide to checking VRAM usage in Windows shows where Task Manager reports dedicated and shared GPU memory.

How to change CUDA - Sysmem Fallback Policy

NVIDIA recommends setting this per program, rather than globally. These steps follow NVIDIA's guide:

  1. Find the program's executable. Start your AI app, open Task Manager and find the process doing the work. For most Stable Diffusion and many local AI front ends, that's python.exe. Right-click it and choose Open file location (or note its path), because many apps ship their own copy of Python.

  2. Open NVIDIA Control Panel. Right-click the desktop and select NVIDIA Control Panel.

  3. Under 3D Settings, click Manage 3D settings.

  4. Open the Program Settings tab and click Add.

  5. Click Browse, select the executable from step 1 and click Open.

  6. Find CUDA - Sysmem Fallback Policy and set it to Prefer No Sysmem Fallback.

  7. Click Apply.

  8. Restart the app. NVIDIA says the app must be restarted for the change to apply.

How to undo it

Go back to Manage 3D settings > Program Settings, select the same executable, set CUDA - Sysmem Fallback Policy to Driver Default, click Apply and restart the app.

If you now get out-of-memory errors

That's expected when a job needs more VRAM than your card has. To make it fit:

  • Lower the image size or batch size in your image generator, or try a smaller model.

  • Use low-VRAM options. Our guide to running ComfyUI with low VRAM covers them.

  • For language models, choose a smaller model or quantisation; see running a local LLM on 8GB of VRAM.

  • Close other GPU-heavy apps, such as browsers and games, before you start.

  • Consider more VRAM for your next upgrade; our how much VRAM guide explains what different workloads need.

Our verdict

NVIDIA's system memory fallback stops CUDA apps crashing when they run out of VRAM by using slower shared system memory instead, which can make Stable Diffusion and other local AI apps suddenly slow. On driver 546.01 or later you can turn it off per app: in NVIDIA Control Panel, add the app's python.exe under Manage 3D settings > Program Settings, set CUDA - Sysmem Fallback Policy to Prefer No Sysmem Fallback, apply and restart the app. Expect out-of-memory errors instead of slowdowns, and set it back to Driver Default to undo. Forge PC hasn't benchmarked the setting.

The bottom line

NVIDIA's system memory fallback stops CUDA apps crashing when they run out of VRAM by using slower shared system memory instead, which can make Stable Diffusion and other local AI apps suddenly slow. On driver 546.01 or later you can turn it off per app: in NVIDIA Control Panel, add the app's python.exe under Manage 3D settings > Program Settings, set CUDA - Sysmem Fallback Policy to Prefer No Sysmem Fallback, apply and restart the app. Expect out-of-memory errors instead of slowdowns, and set it back to Driver Default to undo. Forge PC hasn't benchmarked the setting.

Sources

  1. 01NVIDIA Support, System Memory Fallback for Stable Diffusion

About the author

Forge PC Staff

Editorial team

The shared byline for articles researched, written and edited by the Forge PC editorial team. How we source, check and correct our work is set out in our editorial policy.

More from Forge PC Staff

Read next

Windows Hello Face vs Fingerprint: Which Is Better for a Desktop PC?

Versus

Windows Hello Face vs Fingerprint: Which Is Better for a Desktop PC?

Both work well on a desktop, but you need to add hardware. Face sign-in needs a Windows Hello infrared camera, not a normal webcam, and is the convenient choice if you want a webcam anyway. A USB fingerprint reader is the simpler add-on if you don't want a camera. Microsoft requires a false accept rate below 0.001% for face and large fingerprint sensors.

Forge PC Staff3 min

No Sound on Windows 11? How to Fix Speakers and Headsets

How-To

No Sound on Windows 11? How to Fix Speakers and Headsets

First check Windows is sending audio to the right device: on a desktop, a monitor connected over HDMI or DisplayPort is a common culprit. Then check cables, jacks, mute and volume, turn off audio enhancements, run the Get Help audio troubleshooter, check the volume mixer, install updates and restart, and finally update or reinstall the audio driver.

Forge PC Staff3 min

How to Check Your Motherboard Model in Windows 11

How-To

How to Check Your Motherboard Model in Windows 11

Press Windows + R, type msinfo32 and look for Baseboard Manufacturer and Baseboard Product on the System Summary page. For copyable text, run Get-CimInstance Win32_BaseBoard in PowerShell, or use CPU-Z's Mainboard tab for BIOS and chipset details too. On prebuilts and laptops, use the PC maker's model number and support page instead.

Forge PC Staff3 min

Best Free File Compression Software for Windows in 2026

Best Picks

Best Free File Compression Software for Windows in 2026

7-Zip is the best free archiver for most people: free for any use, it creates encrypted 7z and ZIP files and opens dozens of formats including RAR. NanaZip adds proper Windows 11 context-menu integration and extra formats, and PeaZip supports over 200 archive types. Windows 11 24H2 can open ZIP, RAR, 7z and TAR itself, but not encrypted archives.

Forge PC Staff3 min