github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-08 00:51:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #6997] CUDA error: device kernel image is invalid - CC 7.5 #66486

New Issue
Closed
opened 2026-05-04 06:45:54 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-05-04 06:45:54 -05:00
Owner
Copy Link

Originally created by @nikita228gym on GitHub (Sep 27, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/6997

Originally assigned to: @dhiltgen on GitHub.

What is the issue?

Hello, I would like to apologize for my poor English (I am using a translator from another language). Could you please help me?

I had a problem with Ollama. As always, I tried to run it with the command "Ollama ru llama3.1:8b," but then an error occurred: "Error: llama runner process has terminated: CUDA error: device kernel image is invalid current device: 0, in function ggml_cuda_compute_forward at C:\a\ollama\ollama/llm/llama.cpp/ggml/src/ggml-cuda.cu:2326".

At first, I thought it might be because my version 3.1 was outdated, so I needed to update it. I installed version 3.2, but now, when I try to run 3.2, it gives me this error: Error: lama runner has terminated.

image

my NVIDIA GeForce GTX 1660 Super graphics card
I just talked to Ollama3.1:8b for 2 weeks, and today it stopped working for me

OS

Windows

GPU

Nvidia

CPU

AMD

Ollama version

3.1-3.2

Originally created by @nikita228gym on GitHub (Sep 27, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/6997 Originally assigned to: @dhiltgen on GitHub. ### What is the issue? Hello, I would like to apologize for my poor English (I am using a translator from another language). Could you please help me? I had a problem with Ollama. As always, I tried to run it with the command "Ollama ru llama3.1:8b," but then an error occurred: "Error: llama runner process has terminated: CUDA error: device kernel image is invalid current device: 0, in function ggml_cuda_compute_forward at C:\a\ollama\ollama/llm/llama.cpp/ggml/src/ggml-cuda.cu:2326". At first, I thought it might be because my version 3.1 was outdated, so I needed to update it. I installed version 3.2, but now, when I try to run 3.2, it gives me this error: Error: lama runner has terminated. ![image](https://github.com/user-attachments/assets/e2a9e730-a9fb-42af-972f-6cc4de4453c2) my NVIDIA GeForce GTX 1660 Super graphics card I just talked to Ollama3.1:8b for 2 weeks, and today it stopped working for me ### OS Windows ### GPU Nvidia ### CPU AMD ### Ollama version 3.1-3.2
GiteaMirror added the nvidiabugwindowsneeds more info labels 2026-05-04 06:45:57 -05:00
GiteaMirror commented 2026-05-04 06:46:06 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 27, 2024):

This may be a driver mismatch, other people have fixed similar issues by re-installing (https://github.com/deepchem/deepchem/issues/2156#issuecomment-690843354).

Have you updated the Nvidia driver recently? What's the output of nvidia-smi?

<!-- gh-comment-id:2378514301 --> @rick-github commented on GitHub (Sep 27, 2024): This may be a driver mismatch, other people have fixed similar issues by re-installing (https://github.com/deepchem/deepchem/issues/2156#issuecomment-690843354). Have you updated the Nvidia driver recently? What's the output of `nvidia-smi`?
GiteaMirror commented 2026-05-04 06:46:08 -05:00
Author
Owner
Copy Link

@nikita228gym commented on GitHub (Sep 27, 2024):

This may be a driver mismatch, other people have fixed similar issues by re-installing (deepchem/deepchem#2156 (comment)).

Have you updated the Nvidia driver recently? What's the output of nvidia-smi?

no, I haven't updated the drivers. That's what this command gives you:
image

<!-- gh-comment-id:2378517779 --> @nikita228gym commented on GitHub (Sep 27, 2024): > This may be a driver mismatch, other people have fixed similar issues by re-installing ([deepchem/deepchem#2156 (comment)](https://github.com/deepchem/deepchem/issues/2156#issuecomment-690843354)). > > Have you updated the Nvidia driver recently? What's the output of `nvidia-smi`? no, I haven't updated the drivers. That's what this command gives you: ![image](https://github.com/user-attachments/assets/20657561-e30d-4404-b980-fa1dddb81824)
GiteaMirror commented 2026-05-04 06:46:09 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Sep 28, 2024):

I believe your GPU is a Compute Capability 7.5 GPU, which is one of the architectures we compile both cuda v11 and v12 for.

Did this work on an older version of Ollama? If so, can you try installing that older version and see if it still works?

I believe we should default to the cuda v12 runner for your system. A possible workaround to try would be forcing it to use the cuda v11 runner instead and maybe that works on your GPU. Set OLLAMA_LLM_LIBRARY=cuda_v11

https://github.com/ollama/ollama/blob/main/docs/faq.md#how-do-i-configure-ollama-server

<!-- gh-comment-id:2381020924 --> @dhiltgen commented on GitHub (Sep 28, 2024): I believe your GPU is a Compute Capability 7.5 GPU, which is one of the architectures we compile both cuda v11 and v12 for. Did this work on an older version of Ollama? If so, can you try installing that older version and see if it still works? I believe we should default to the cuda v12 runner for your system. A possible workaround to try would be forcing it to use the cuda v11 runner instead and maybe that works on your GPU. Set OLLAMA_LLM_LIBRARY=cuda_v11 https://github.com/ollama/ollama/blob/main/docs/faq.md#how-do-i-configure-ollama-server
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 needs more info nvidia windows
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#66486
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?