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 #6647] can't use nvidia GPU only after sleep #29943

New Issue
Closed
opened 2026-04-22 09:18:09 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-04-22 09:18:09 -05:00
Owner
Copy Link

Originally created by @brookate on GitHub (Sep 5, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/6647

What is the issue?

after gentoo linux sleep, ollama only use cpu
turn on OOLAMA_DEBUG, I find such line

time=2024-09-05T09:20:35.622+08:00 level=DEBUG source=gpu.go:521 msg="discovered GPU libraries" paths="[/tmp/ollama786265597/runners/cuda_v11/libcudart.so.11.0 /opt/cuda/lib64/libcudart.so.12.6.37]"
cudaSetDevice err: 999
time=2024-09-05T09:20:35.624+08:00 level=DEBUG source=gpu.go:533 msg="Unable to load cudart" library=/tmp/ollama786265597/runners/cuda_v11/libcudart.so.11.0 error="cudart init failure: 999"
cudaSetDevice err: 999
time=2024-09-05T09:20:35.627+08:00 level=DEBUG source=gpu.go:533 msg="Unable to load cudart" library=/opt/cuda/lib64/libcudart.so.12.6.37 error="cudart init failure: 999"

my cuda is
dell➜ ~ cd /opt/cuda
dell➜ cuda l
total 88K
drwxr-xr-x 6 root root 4.0K Aug 28 10:01 .
drwxr-xr-x 14 root root 4.0K Aug 26 18:22 ..
-rw-r--r-- 1 root root 62K Aug 28 10:00 EULA.txt
drwxr-xr-x 3 root root 4.0K Aug 28 10:01 bin
lrwxrwxrwx 1 root root 28 Aug 28 10:00 include -> targets/x86_64-linux/include
lrwxrwxrwx 1 root root 24 Aug 28 10:00 lib64 -> targets/x86_64-linux/lib
drwxr-xr-x 3 root root 4.0K Jul 31 18:53 nvml
drwxr-xr-x 6 root root 4.0K Jul 31 18:53 nvvm
drwxr-xr-x 3 root root 4.0K Jul 31 18:53 targets
dell➜ cuda cd targets/x86_64-linux/lib
dell➜ lib ls libcudart.so
libcudart.so@ libcudart.so.12@ libcudart.so.12.6.37*

OS

Linux

GPU

Nvidia

CPU

Intel

Ollama version

0.3.0

Originally created by @brookate on GitHub (Sep 5, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/6647 ### What is the issue? after gentoo linux sleep, ollama only use cpu turn on OOLAMA_DEBUG, I find such line ``` time=2024-09-05T09:20:35.622+08:00 level=DEBUG source=gpu.go:521 msg="discovered GPU libraries" paths="[/tmp/ollama786265597/runners/cuda_v11/libcudart.so.11.0 /opt/cuda/lib64/libcudart.so.12.6.37]" cudaSetDevice err: 999 time=2024-09-05T09:20:35.624+08:00 level=DEBUG source=gpu.go:533 msg="Unable to load cudart" library=/tmp/ollama786265597/runners/cuda_v11/libcudart.so.11.0 error="cudart init failure: 999" cudaSetDevice err: 999 time=2024-09-05T09:20:35.627+08:00 level=DEBUG source=gpu.go:533 msg="Unable to load cudart" library=/opt/cuda/lib64/libcudart.so.12.6.37 error="cudart init failure: 999" ``` my cuda is dell➜ ~ cd /opt/cuda dell➜ cuda l total 88K drwxr-xr-x 6 root root 4.0K Aug 28 10:01 . drwxr-xr-x 14 root root 4.0K Aug 26 18:22 .. -rw-r--r-- 1 root root 62K Aug 28 10:00 EULA.txt drwxr-xr-x 3 root root 4.0K Aug 28 10:01 bin lrwxrwxrwx 1 root root 28 Aug 28 10:00 include -> targets/x86_64-linux/include lrwxrwxrwx 1 root root 24 Aug 28 10:00 lib64 -> targets/x86_64-linux/lib drwxr-xr-x 3 root root 4.0K Jul 31 18:53 nvml drwxr-xr-x 6 root root 4.0K Jul 31 18:53 nvvm drwxr-xr-x 3 root root 4.0K Jul 31 18:53 targets dell➜ cuda cd targets/x86_64-linux/lib dell➜ lib ls libcudart.so libcudart.so@ libcudart.so.12@ libcudart.so.12.6.37* ### OS Linux ### GPU Nvidia ### CPU Intel ### Ollama version 0.3.0
GiteaMirror added the bug label 2026-04-22 09:18:09 -05:00
GiteaMirror commented 2026-04-22 09:18:10 -05:00
Author
Owner
Copy Link

@brookate commented on GitHub (Sep 5, 2024):

Ok, I find a way after some search.

sudo rmmod nvidia_uvm
sudo modprobe nvidia_uvm

it works to me.

<!-- gh-comment-id:2330433043 --> @brookate commented on GitHub (Sep 5, 2024): Ok, I find a way after some search. ``` sudo rmmod nvidia_uvm sudo modprobe nvidia_uvm ``` it works to me.
GiteaMirror commented 2026-04-22 09:18:11 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Sep 5, 2024):

Poor suspend/resume behavior on nvidia GPUs is tracked via #5464

<!-- gh-comment-id:2332156618 --> @dhiltgen commented on GitHub (Sep 5, 2024): Poor suspend/resume behavior on nvidia GPUs is tracked via #5464
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#29943
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?