github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-08 00:51:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #13622] poor vram allocations #34724

New Issue
Closed
opened 2026-04-22 18:32:51 -05:00 by GiteaMirror · 1 comment
GiteaMirror commented 2026-04-22 18:32:51 -05:00
Owner
Copy Link

Originally created by @wOvAN on GitHub (Jan 4, 2026).
Original GitHub issue: https://github.com/ollama/ollama/issues/13622

What is the issue?

Image as you see, a lot of vram not used at all, while model gets offloaded to ram. for example with llama.cpp I can run larger context and quants on the same system.

also would be good to load a model layers in parallel for multi gpus at once in cases where model storage faster than gpu's pci link speed.

Relevant log output

OS

Linux, Docker

GPU

Nvidia

CPU

No response

Ollama version

0.13.5

Originally created by @wOvAN on GitHub (Jan 4, 2026). Original GitHub issue: https://github.com/ollama/ollama/issues/13622 ### What is the issue? 1. <img width="1650" height="496" alt="Image" src="https://github.com/user-attachments/assets/b820ff14-9f50-4062-905b-cca2a0f21103" /> as you see, a lot of vram not used at all, while model gets offloaded to ram. for example with llama.cpp I can run larger context and quants on the same system. 2. also would be good to load a model layers in parallel for multi gpus at once in cases where model storage faster than gpu's pci link speed. ### Relevant log output ```shell ``` ### OS Linux, Docker ### GPU Nvidia ### CPU _No response_ ### Ollama version 0.13.5
GiteaMirror added the bug label 2026-04-22 18:32:51 -05:00
GiteaMirror commented 2026-04-22 18:32:53 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Jan 5, 2026):

GLM is not yet supported in the new ollama engine and so uses the older memory estimation logic, which is sometimes inaccurate. More VRAM can be used by explictly setting num_gpu.

<!-- gh-comment-id:3708575766 --> @rick-github commented on GitHub (Jan 5, 2026): GLM is not yet supported in the new ollama engine and so uses the older memory estimation logic, which is sometimes inaccurate. More VRAM can be used by explictly setting [`num_gpu`](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#34724
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?