github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-11 11:07:38 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #3177] GPU utilization & Context Length and Max Tokens & Command-line windows crash & Server connection failed #1957

New Issue
Closed
opened 2026-04-12 12:06:24 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-04-12 12:06:24 -05:00
Owner
Copy Link

Originally created by @HWiwoiiii on GitHub (Mar 16, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/3177

Originally assigned to: @dhiltgen on GitHub.

What is the issue?

  1. How to setup Ollama for models to use my GPU?
    I'm using Windows with a 32GB DDR4 2667MHz memory (16GB + 16GB) and an NVIDIA GeForce RTX 2080 Super with Max-Q Design (8GB / Dell). Intel(R) UHD Graphics (1GB / Dell). However, the Ollama doesn't seem to utilize my GPU even when I have both my CPU and memory running at 99% utilization. Is this a bug or are there any settings that I am accidentally ignoring?

  2. Does adjusting Context Length and Max Tokens have any impact on running the model?
    Are the Context Length and Max Tokens predefined or that we can change them in open webui? Can I extend myh input and the outputs by adjusting them? I know this is a silly question so if anyone is tired explaining it leaving a link about that here would be greatly appreciated as well.

  3. How to keep Ollama running?
    The cmd window that I use to run a model always crash and disappears when I want to type something in it is it normal? By the way do I need to run ollama run gemma:7b every time I restart my pc?

image
image
image

What did you expect to see?

  1. How to setup Ollama for models to use my GPU?
    GPU being used

  2. Does adjusting Context Length and Max Tokens have any impact on running the model?

  3. How to keep Ollama running?
    Ollama running correctly and can be accessed by open webui
    image

Steps to reproduce

restart my pc

Are there any recent changes that introduced the issue?

not working

OS

Windows

Architecture

Other

Platform

No response

Ollama version

latest

GPU

Intel

GPU info

NVIDIA GeForce RTX 2080 Super with Max-Q Design (8GB / Dell). Intel(R) UHD Graphics (1GB / Dell).

CPU

Intel

Other software

open webui

Originally created by @HWiwoiiii on GitHub (Mar 16, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/3177 Originally assigned to: @dhiltgen on GitHub. ### What is the issue? 1. How to setup Ollama for models to use my GPU? I'm using Windows with a 32GB DDR4 2667MHz memory (16GB + 16GB) and an NVIDIA GeForce RTX 2080 Super with Max-Q Design (8GB / Dell). Intel(R) UHD Graphics (1GB / Dell). However, the Ollama doesn't seem to utilize my GPU even when I have both my CPU and memory running at 99% utilization. Is this a bug or are there any settings that I am accidentally ignoring? 2. Does adjusting Context Length and Max Tokens have any impact on running the model? Are the Context Length and Max Tokens predefined or that we can change them in open webui? Can I extend myh input and the outputs by adjusting them? I know this is a silly question so if anyone is tired explaining it leaving a link about that here would be greatly appreciated as well. 3. How to keep Ollama running? The cmd window that I use to run a model always crash and disappears when I want to type something in it is it normal? By the way do I need to run `ollama run gemma:7b` every time I restart my pc? ![image](https://github.com/ollama/ollama/assets/103039908/764adea7-283b-4a05-ace0-92763d0f6028) ![image](https://github.com/ollama/ollama/assets/103039908/d219de7a-c5ff-44e1-af2c-90938d83889d) ![image](https://github.com/ollama/ollama/assets/103039908/8c3c1296-3be6-4862-ab23-45264d95427f) ### What did you expect to see? 1. How to setup Ollama for models to use my GPU? GPU being used 2. Does adjusting Context Length and Max Tokens have any impact on running the model? 3. How to keep Ollama running? Ollama running correctly and can be accessed by open webui ![image](https://github.com/ollama/ollama/assets/103039908/179ccb1b-9801-4a31-90fd-8573690f9989) ### Steps to reproduce restart my pc ### Are there any recent changes that introduced the issue? not working ### OS Windows ### Architecture Other ### Platform _No response_ ### Ollama version latest ### GPU Intel ### GPU info NVIDIA GeForce RTX 2080 Super with Max-Q Design (8GB / Dell). Intel(R) UHD Graphics (1GB / Dell). ### CPU Intel ### Other software open webui
GiteaMirror added the bugnvidiawindows labels 2026-04-12 12:06:25 -05:00
GiteaMirror commented 2026-04-12 12:06:27 -05:00
Author
Owner
Copy Link

@HWiwoiiii commented on GitHub (Mar 16, 2024):

Update:
I restarted my PC, and initially, the Ollama works fine for the first few seconds. However, when I ask something else, it gives the error shown below.
image
image

Another thing is, In the task manager, the Vmmem process cannot be ended. I understand that it is related to virtual memory and other system functions. However, even if I close both Ollama and Docker, the process persists, it doesn't seem to be directly linked to either Ollama or Docker. I just want to confirm if it is a regular Windows process, as it cannot be terminated. Is it related to Ollama?
image

<!-- gh-comment-id:2001404802 --> @HWiwoiiii commented on GitHub (Mar 16, 2024): Update: I restarted my PC, and initially, the Ollama works fine for the first few seconds. However, when I ask something else, it gives the error shown below. ![image](https://github.com/ollama/ollama/assets/103039908/87e670cb-8072-4a58-aed1-c8b339f6ad8c) ![image](https://github.com/ollama/ollama/assets/103039908/3b1371ae-02e3-4235-bfc0-11e66c94bd35) Another thing is, In the task manager, the Vmmem process cannot be ended. I understand that it is related to virtual memory and other system functions. However, even if I close both Ollama and Docker, the process persists, it doesn't seem to be directly linked to either Ollama or Docker. I just want to confirm if it is a regular Windows process, as it cannot be terminated. Is it related to Ollama? ![image](https://github.com/ollama/ollama/assets/103039908/6173a666-cbb6-42ed-9db7-bc65be957dcd)
GiteaMirror commented 2026-04-12 12:06:27 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Mar 20, 2024):

Can you share your server log? It sounds like the server is probably crashing for some reason, and we'll be able to understand why from the log. This will also help us understand if it is correctly identifying your GPU.

<!-- gh-comment-id:2009082908 --> @dhiltgen commented on GitHub (Mar 20, 2024): Can you share your server log? It sounds like the server is probably crashing for some reason, and we'll be able to understand why from the log. This will also help us understand if it is correctly identifying your GPU.
GiteaMirror commented 2026-04-12 12:06:28 -05:00
Author
Owner
Copy Link

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

We've fixed a number of stability bugs on Windows in 0.1.33 (available in pre-release). Please give the update a try and if you're still seeing problems, please share your server log and I'll re-open the issue to investigate.

<!-- gh-comment-id:2081605376 --> @dhiltgen commented on GitHub (Apr 28, 2024): We've fixed a number of stability bugs on Windows in [0.1.33](https://github.com/ollama/ollama/releases) (available in pre-release). Please give the update a try and if you're still seeing problems, please share your server log and I'll re-open the issue to investigate.
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#1957
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?