github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 16:40:08 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #4500] High CPU Usage and Model Stoppage Issue in Ollama on Linux CentOS7 Without GPU #28578

New Issue
Closed
opened 2026-04-22 06:52:52 -05:00 by GiteaMirror · 1 comment
GiteaMirror commented 2026-04-22 06:52:52 -05:00
Owner
Copy Link

Originally created by @sirfuwh on GitHub (May 17, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/4500

What is the issue?

I'm using Ollama as the framework for large AI models, with open-webUI or anythingLLM as the frontend.
My machine is running Linux CentOS7 with 32GB of memory and a 24-core CPU, but no GPU.
When running models like phi3 or others in Ollama, the CPU usage is around 1200% (htop shows 12 cores fully utilized). After a few rounds of conversation, which might take about 5-6 minutes, the CPU usage drops to 100% (htop shows one random core fully utilized), and it stays that way until I stop Ollama. At this point, the model stops generating text. Not only does the current model stop working, but switching to other models downloaded in Ollama also has no effect, although the Linux system itself does not crash.
It seems to be an issue with Ollama. And dmsg and journalctl -u ollama no special hits .
ollama_100percent_usage
thanks!

OS

Linux

GPU

No response

CPU

Intel

Ollama version

0.1.30

Originally created by @sirfuwh on GitHub (May 17, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/4500 ### What is the issue? I'm using Ollama as the framework for large AI models, with open-webUI or anythingLLM as the frontend. My machine is running Linux CentOS7 with 32GB of memory and a 24-core CPU, but no GPU. When running models like phi3 or others in Ollama, the CPU usage is around 1200% (htop shows 12 cores fully utilized). After a few rounds of conversation, which might take about 5-6 minutes, the CPU usage drops to 100% (htop shows one random core fully utilized), and it stays that way until I stop Ollama. At this point, the model stops generating text. Not only does the current model stop working, but switching to other models downloaded in Ollama also has no effect, although the Linux system itself does not crash. It seems to be an issue with Ollama. And `dmsg` and `journalctl -u ollama` no special hits . ![ollama_100percent_usage](https://github.com/ollama/ollama/assets/58595497/1aca8c62-232c-4d10-9ea2-da58bfeeb12f) thanks! ### OS Linux ### GPU _No response_ ### CPU Intel ### Ollama version 0.1.30
GiteaMirror added the bug label 2026-04-22 06:52:52 -05:00
GiteaMirror commented 2026-04-22 06:52:54 -05:00
Author
Owner
Copy Link

@sirfuwh commented on GitHub (May 17, 2024):

After I update to latest 0.1.38 verson . No ploblem now.
And When testing , I find Anythin LLM runing with 25.G ! maybe this is also a factor.

<!-- gh-comment-id:2117851729 --> @sirfuwh commented on GitHub (May 17, 2024): After I update to latest 0.1.38 verson . No ploblem now. And When testing , I find Anythin LLM runing with 25.G ! maybe this is also a factor.
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#28578
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?