github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 16:11:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #7721] AMD graphics card encounters a memory usage exception while running on Windows 11. #51440

New Issue
Closed
opened 2026-04-28 20:08:26 -05:00 by GiteaMirror · 4 comments
GiteaMirror commented 2026-04-28 20:08:26 -05:00
Owner
Copy Link

Originally created by @uamind on GitHub (Nov 18, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/7721

What is the issue?

cpu: ryzen 3900xt
gpu: amd 6800xt
memory: ddr4 64g
os: windows 11 23h2
ui: open-webui
model: mistral-nemo

An scheduling exception related to GPU memory occurred during the execution of Mistral-Nemo, causing the GPU memory usage to be allocated to shared GPU memory instead of the actual GPU memory (even though there was sufficient GPU memory available), resulting in a total usage of 11GB. However, there were still 11.6GB of GPU memory left unused. This resulted in slow performance with only 3.89 tokens being processed.

Previously, running this same 12B model did not allocate shared GPU memory, but when using the Mistral model, it could utilize up to 52.28 tokens with a GPU memory usage of 8.4GB.

Screenshot 2024-11-18 152716

OS

Windows

GPU

AMD

CPU

AMD

Ollama version

0.4.2

Originally created by @uamind on GitHub (Nov 18, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/7721 ### What is the issue? cpu: ryzen 3900xt gpu: amd 6800xt memory: ddr4 64g os: windows 11 23h2 ui: open-webui model: mistral-nemo An scheduling exception related to GPU memory occurred during the execution of Mistral-Nemo, causing the GPU memory usage to be allocated to shared GPU memory instead of the actual GPU memory (even though there was sufficient GPU memory available), resulting in a total usage of 11GB. However, there were still 11.6GB of GPU memory left unused. This resulted in slow performance with only 3.89 tokens being processed. Previously, running this same 12B model did not allocate shared GPU memory, but when using the Mistral model, it could utilize up to 52.28 tokens with a GPU memory usage of 8.4GB. ![Screenshot 2024-11-18 152716](https://github.com/user-attachments/assets/b46425d7-667f-405c-98cc-dc5a6adc6ac4) ### OS Windows ### GPU AMD ### CPU AMD ### Ollama version 0.4.2
GiteaMirror added the bug label 2026-04-28 20:08:26 -05:00
GiteaMirror commented 2026-04-28 20:08:32 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Nov 18, 2024):

Server logs may aid in debugging.

<!-- gh-comment-id:2482552430 --> @rick-github commented on GitHub (Nov 18, 2024): [Server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues) may aid in debugging.
GiteaMirror commented 2026-04-28 20:08:37 -05:00
Author
Owner
Copy Link

@galizhur commented on GitHub (Nov 18, 2024):

It looks like we have the same issue https://github.com/ollama/ollama/issues/7715

<!-- gh-comment-id:2482804616 --> @galizhur commented on GitHub (Nov 18, 2024): It looks like we have the same issue https://github.com/ollama/ollama/issues/7715
GiteaMirror commented 2026-04-28 20:08:40 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Nov 18, 2024):

This is a known driver bug tracked in #7107. Until the next driver comes out, the workaround is to downgrade to an older driver.

<!-- gh-comment-id:2484224517 --> @dhiltgen commented on GitHub (Nov 18, 2024): This is a known driver bug tracked in #7107. Until the next driver comes out, the workaround is to downgrade to an older driver.
GiteaMirror commented 2026-04-28 20:08:43 -05:00
Author
Owner
Copy Link

@uamind commented on GitHub (Nov 19, 2024):

Thank you.

<!-- gh-comment-id:2484866030 --> @uamind commented on GitHub (Nov 19, 2024): Thank you.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#51440
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?