github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-23 22:21:55 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #10884] Help: Ollama Only Uses 1.8/4.0 GB VRAM, Swapping 2.5 GB to RAM #53665

New Issue
Closed
opened 2026-04-29 04:26:10 -05:00 by GiteaMirror · 1 comment
GiteaMirror commented 2026-04-29 04:26:10 -05:00
Owner
Copy Link

Originally created by @TDuy22 on GitHub (May 28, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/10884

Help: Ollama Only Uses 1.8/4.0 GB VRAM, Swapping 2.5 GB to RAM

Hi everyone,

I'm running a local LLM using Ollama on my machine, but I'm encountering an issue where it only utilizes 1.8 GB out of the available 4.0 GB VRAM on my GPU, with 2.5 GB being swapped to RAM. This is causing slower computation speeds (my gpu only run about 30%), and I’d like to maximize VRAM usage to improve performance. I’d appreciate any insights or solutions from the community!
System and Model Details

GPU: NVIDIA RTX 3050 laptop (4 GB VRAM)

Model: gemma3:4b (3.3gb)
Architecture: gemma3
Parameters: 4.3B
Context Length: 131072 (but set to num_ctx=1024)
Embedding Length: 2560
Quantization: Q4_K_M
Capabilities: Completion, Vision
Parameters:
num_ctx: 1024
stop: "<end_of_turn>"
temperature: 1
top_k: 64
top_p: 0.95
License: Gemma Terms of Use (Last modified: February 21, 2024)
Model Size: Approximately 3.3 GB
Ollama Version: 0.7.1

Relevant log output

OS

Windows

GPU

Nvidia

CPU

AMD

Ollama version

0.7.1

Originally created by @TDuy22 on GitHub (May 28, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/10884 ### Help: Ollama Only Uses 1.8/4.0 GB VRAM, Swapping 2.5 GB to RAM Hi everyone, I'm running a local LLM using Ollama on my machine, but I'm encountering an issue where it only utilizes 1.8 GB out of the available 4.0 GB VRAM on my GPU, with 2.5 GB being swapped to RAM. This is causing slower computation speeds (my gpu only run about 30%), and I’d like to maximize VRAM usage to improve performance. I’d appreciate any insights or solutions from the community! System and Model Details GPU: NVIDIA RTX 3050 laptop (4 GB VRAM) Model: gemma3:4b (3.3gb) Architecture: gemma3 Parameters: 4.3B Context Length: 131072 (but set to num_ctx=1024) Embedding Length: 2560 Quantization: Q4_K_M Capabilities: Completion, Vision Parameters: num_ctx: 1024 stop: "<end_of_turn>" temperature: 1 top_k: 64 top_p: 0.95 License: Gemma Terms of Use (Last modified: February 21, 2024) Model Size: Approximately 3.3 GB Ollama Version: 0.7.1 ### Relevant log output ```shell ``` ### OS Windows ### GPU Nvidia ### CPU AMD ### Ollama version 0.7.1
GiteaMirror added the bug label 2026-04-29 04:26:10 -05:00
GiteaMirror commented 2026-04-29 04:26:13 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (May 28, 2025):

Server logs will give more information, but the likely cause is that ollama is not correctly estimating the amount of layers it can offload. You can override this be setting num_gpu as described here.

<!-- gh-comment-id:2916094586 --> @rick-github commented on GitHub (May 28, 2025): [Server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues) will give more information, but the likely cause is that ollama is not correctly estimating the amount of layers it can offload. You can override this be setting `num_gpu` as described [here](https://github.com/ollama/ollama/issues/6950#issuecomment-2373663650).
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#53665
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?