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 #1867] ollama barely uses any Ram #1067

New Issue
Closed
opened 2026-04-12 10:49:18 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-04-12 10:49:18 -05:00
Owner
Copy Link

Originally created by @neuleo on GitHub (Jan 9, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/1867

Hey Guys,

I run ollama on docker and use mostly 7b models. But my Ram usage stays under 4 GB. Sometimes even below 3 GB.

But the recommendations are 8 GB of Ram.

It has 4 Core CPU, and it generates very slow even though I got 24 GB of Ram. I don't have a Video Card, though.

I'm new to this, so can anyone tell me what I might need to do differently?

Originally created by @neuleo on GitHub (Jan 9, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/1867 Hey Guys, I run ollama on docker and use mostly 7b models. But my Ram usage stays under 4 GB. Sometimes even below 3 GB. But the recommendations are 8 GB of Ram. It has 4 Core CPU, and it generates very slow even though I got 24 GB of Ram. I don't have a Video Card, though. I'm new to this, so can anyone tell me what I might need to do differently?
GiteaMirror commented 2026-04-12 10:49:18 -05:00
Author
Owner
Copy Link

@easp commented on GitHub (Jan 9, 2024):

Models are loaded using mmap and as a result probably appear in file cache memory use, rather than as part of the ollama process memory.

<!-- gh-comment-id:1883432103 --> @easp commented on GitHub (Jan 9, 2024): Models are loaded using mmap and as a result probably appear in file cache memory use, rather than as part of the ollama process memory.
GiteaMirror commented 2026-04-12 10:49:19 -05:00
Author
Owner
Copy Link

@pdevine commented on GitHub (Jan 9, 2024):

@neuleo for a 7b, 4bit quantized model I would expect it to take up around 4 GB. The amount of memory though comes down to the size of the model and the context size that you're using, so it's a bit squishy. We're adding some improvements in 0.1.19 to be able to more accurately guess the amount of memory though.

That said, I don't know what CPU you're using, but generally speaking, you'll get far better results from a GPU than the CPU. We've also got some changes coming to take more advantage of the AVX capabilities in the CPU which if you have a modern CPU w/ AVX-512 you may see some performance gains.

I'm going to go ahead and close the issue, but feel free to keep commenting or reach out on the discord.

<!-- gh-comment-id:1883592927 --> @pdevine commented on GitHub (Jan 9, 2024): @neuleo for a 7b, 4bit quantized model I would expect it to take up around 4 GB. The amount of memory though comes down to the size of the model _and_ the context size that you're using, so it's a bit squishy. We're adding some improvements in 0.1.19 to be able to more accurately guess the amount of memory though. That said, I don't know what CPU you're using, but generally speaking, you'll get far better results from a GPU than the CPU. We've also got some changes coming to take more advantage of the AVX capabilities in the CPU which if you have a modern CPU w/ AVX-512 you may see some performance gains. I'm going to go ahead and close the issue, but feel free to keep commenting or reach out on the discord.
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
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#1067
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?