github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-05 23:53:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #3486] ollama not using GPU in windows while all layers offloaded to gpu #2147

New Issue
Closed
opened 2026-04-12 12:22:41 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-04-12 12:22:41 -05:00
Owner
Copy Link

Originally created by @VSR2007 on GitHub (Apr 4, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/3486

Originally assigned to: @dhiltgen on GitHub.

What is the issue?

new 1.txt
I running ollama windows. I have nvidia rtx 2000 ada generation gpu with 8gb ram. It also have 20 cores cpu with 64gb ram. Ollama some how does not use gpu for inferencing. the GPU shoots up when given a prompt for a moment (<1 s) and then stays at 0/1 %. All this while it occupies only 4.5gb of gpu ram. I am using mistral 7b.

What did you expect to see?

better inference speed with full utilization of gpu especially when gpu ram is not limiting.

Steps to reproduce

Not sure

Are there any recent changes that introduced the issue?

No response

OS

Windows

Architecture

x86

Platform

No response

Ollama version

No response

GPU

Nvidia

GPU info

tt

CPU

Intel

Other software

No response

Originally created by @VSR2007 on GitHub (Apr 4, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/3486 Originally assigned to: @dhiltgen on GitHub. ### What is the issue? [new 1.txt](https://github.com/ollama/ollama/files/14865468/new.1.txt) I running ollama windows. I have nvidia rtx 2000 ada generation gpu with 8gb ram. It also have 20 cores cpu with 64gb ram. Ollama some how does not use gpu for inferencing. the GPU shoots up when given a prompt for a moment (<1 s) and then stays at 0/1 %. All this while it occupies only 4.5gb of gpu ram. I am using mistral 7b. ### What did you expect to see? better inference speed with full utilization of gpu especially when gpu ram is not limiting. ### Steps to reproduce Not sure ### Are there any recent changes that introduced the issue? _No response_ ### OS Windows ### Architecture x86 ### Platform _No response_ ### Ollama version _No response_ ### GPU Nvidia ### GPU info ![tt](https://github.com/ollama/ollama/assets/107546824/2a5aae48-65aa-4e93-a4a5-39fe879a4be2) ### CPU Intel ### Other software _No response_
GiteaMirror added the bugnvidiawindows labels 2026-04-12 12:22:41 -05:00
GiteaMirror commented 2026-04-12 12:22:42 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Apr 12, 2024):

The log you shared indicates it was loading into the GPU. What sort of token rate are you seeing?

I've seen other people report that if they stop the tray application, and run the server in a terminal manually they see much better performance. We still don't understand why, but that might be an experiment to try. Also, can you look at Task Manager when you're seeing the slow performance and let us know if you see anything notable? (e.g., some AV product chewing up a lot of CPU when you run inference, or something that might explain why it's being throttled.)

<!-- gh-comment-id:2052698900 --> @dhiltgen commented on GitHub (Apr 12, 2024): The log you shared indicates it was loading into the GPU. What sort of token rate are you seeing? I've seen other people report that if they stop the tray application, and run the server in a terminal manually they see much better performance. We still don't understand why, but that might be an experiment to try. Also, can you look at Task Manager when you're seeing the slow performance and let us know if you see anything notable? (e.g., some AV product chewing up a lot of CPU when you run inference, or something that might explain why it's being throttled.)
GiteaMirror commented 2026-04-12 12:22:42 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (May 4, 2024):

If you're still having problems and it doesn't match #3511 please share more information and I'll re-open the ticket.

<!-- gh-comment-id:2094422124 --> @dhiltgen commented on GitHub (May 4, 2024): If you're still having problems and it doesn't match #3511 please share more information and I'll re-open the ticket.
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 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#2147
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?