github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 00:22:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #10470] cuda driver library init failure with new nvidia driver version. #32645

New Issue
Closed
opened 2026-04-22 14:17:51 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-04-22 14:17:51 -05:00
Owner
Copy Link

Originally created by @Enlux on GitHub (Apr 29, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/10470

What is the issue?

Hello,

I've recently updated my nvidia drivers on the host system to 570.133.20 and now i'm getting the driver lib init failure below. I've tested nvidia-smi by running docker run --rm -it --gpus all -v ollama:/root/.ollama --name ollama-debug --entrypoint bash ollama/ollama and it shows all gpus.

Relevant log output

2025/04/29 10:30:05 routes.go:1232: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: HTTPS_PROXY: HTTP_PROXY: NO_PROXY: OLLAMA_CONTEXT_LENGTH:2048 OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/root/.ollama/models OLLAMA_MULTIUSER_CACHE:false OLLAMA_NEW_ENGINE:false OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:0 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://* vscode-webview://* vscode-file://*] OLLAMA_SCHED_SPREAD:false ROCR_VISIBLE_DEVICES: http_proxy: https_proxy: no_proxy:]"
time=2025-04-29T10:30:05.206Z level=INFO source=images.go:458 msg="total blobs: 27"
time=2025-04-29T10:30:05.207Z level=INFO source=images.go:465 msg="total unused blobs removed: 0"
time=2025-04-29T10:30:05.207Z level=INFO source=routes.go:1299 msg="Listening on [::]:11434 (version 0.6.6)"
time=2025-04-29T10:30:05.207Z level=INFO source=gpu.go:217 msg="looking for compatible GPUs"
time=2025-04-29T10:30:05.622Z level=INFO source=gpu.go:612 msg="Unable to load cudart library /usr/lib/x86_64-linux-gnu/libcuda.so.570.133.20: cuda driver library init failure: 3"
time=2025-04-29T10:30:06.368Z level=INFO source=gpu.go:377 msg="no compatible GPUs were discovered"
time=2025-04-29T10:30:06.368Z level=INFO source=types.go:130 msg="inference compute" id=0 library=cpu variant="" compute="" driver=0.0 name="" total="864.6 GiB" available="856.5 GiB"

OS

Linux, Docker

GPU

Nvidia

CPU

AMD

Ollama version

0.6.6

Originally created by @Enlux on GitHub (Apr 29, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/10470 ### What is the issue? Hello, I've recently updated my nvidia drivers on the host system to 570.133.20 and now i'm getting the driver lib init failure below. I've tested nvidia-smi by running `docker run --rm -it --gpus all -v ollama:/root/.ollama --name ollama-debug --entrypoint bash ollama/ollama` and it shows all gpus. ### Relevant log output ```shell 2025/04/29 10:30:05 routes.go:1232: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: HTTPS_PROXY: HTTP_PROXY: NO_PROXY: OLLAMA_CONTEXT_LENGTH:2048 OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/root/.ollama/models OLLAMA_MULTIUSER_CACHE:false OLLAMA_NEW_ENGINE:false OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:0 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://* vscode-webview://* vscode-file://*] OLLAMA_SCHED_SPREAD:false ROCR_VISIBLE_DEVICES: http_proxy: https_proxy: no_proxy:]" time=2025-04-29T10:30:05.206Z level=INFO source=images.go:458 msg="total blobs: 27" time=2025-04-29T10:30:05.207Z level=INFO source=images.go:465 msg="total unused blobs removed: 0" time=2025-04-29T10:30:05.207Z level=INFO source=routes.go:1299 msg="Listening on [::]:11434 (version 0.6.6)" time=2025-04-29T10:30:05.207Z level=INFO source=gpu.go:217 msg="looking for compatible GPUs" time=2025-04-29T10:30:05.622Z level=INFO source=gpu.go:612 msg="Unable to load cudart library /usr/lib/x86_64-linux-gnu/libcuda.so.570.133.20: cuda driver library init failure: 3" time=2025-04-29T10:30:06.368Z level=INFO source=gpu.go:377 msg="no compatible GPUs were discovered" time=2025-04-29T10:30:06.368Z level=INFO source=types.go:130 msg="inference compute" id=0 library=cpu variant="" compute="" driver=0.0 name="" total="864.6 GiB" available="856.5 GiB" ``` ### OS Linux, Docker ### GPU Nvidia ### CPU AMD ### Ollama version 0.6.6
GiteaMirror added the bug label 2026-04-22 14:17:51 -05:00
GiteaMirror commented 2026-04-22 14:17:55 -05:00
Author
Owner
Copy Link

@Enlux commented on GitHub (Apr 30, 2025):

The issue wasn't related to ollama but rather to gpu passthrough and nvidia driver version.

<!-- gh-comment-id:2841640720 --> @Enlux commented on GitHub (Apr 30, 2025): The issue wasn't related to ollama but rather to gpu passthrough and nvidia driver version.
GiteaMirror commented 2026-04-22 14:17:57 -05:00
Author
Owner
Copy Link

@vladrad commented on GitHub (May 16, 2025):

hi I am having the same problem... what was your solution?

<!-- gh-comment-id:2887784961 --> @vladrad commented on GitHub (May 16, 2025): hi I am having the same problem... what was your solution?
GiteaMirror commented 2026-04-22 14:17:58 -05:00
Author
Owner
Copy Link

@Enlux commented on GitHub (May 17, 2025):

hi I am having the same problem... what was your solution?

Hey, I was using proxmox and was running this in a vm while passing through 5 gpus. Ever since i upgraded my server to 1TB of ram from 512gb the whole passthrough has been buggy. If I tried to use 4/5 gpus for some reason nothing worked and only when I used 3 or less gpus (no matter which ones) cuda started working. nvidia-smi was showing all 5 gpus though. After figuring out it was somehow related to virtualization, I decided to ditch the proxmox/virtualization layer completely and booted directly into Ubuntu instead. Everything works as intended now.
Usually this error code means something more simple is happening e.g. cuda container toolkit not being installed

<!-- gh-comment-id:2888368209 --> @Enlux commented on GitHub (May 17, 2025): > hi I am having the same problem... what was your solution? Hey, I was using proxmox and was running this in a vm while passing through 5 gpus. Ever since i upgraded my server to 1TB of ram from 512gb the whole passthrough has been buggy. If I tried to use 4/5 gpus for some reason nothing worked and only when I used 3 or less gpus (no matter which ones) cuda started working. nvidia-smi was showing all 5 gpus though. After figuring out it was somehow related to virtualization, I decided to ditch the proxmox/virtualization layer completely and booted directly into Ubuntu instead. Everything works as intended now. Usually this error code means something more simple is happening e.g. cuda container toolkit not being installed
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#32645
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?