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 #12328] Ollama is still broken for Thor and Spark #70249

New Issue
Closed
opened 2026-05-04 20:46:46 -05:00 by GiteaMirror · 6 comments
GiteaMirror commented 2026-05-04 20:46:46 -05:00
Owner
Copy Link

Originally created by @johnnynunez on GitHub (Sep 18, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/12328

What is the issue?

Not taking good CUDA gpu execution. 22% as maximum

cc @dhiltgen @mxyng

Originally created by @johnnynunez on GitHub (Sep 18, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/12328 ### What is the issue? Not taking good CUDA gpu execution. 22% as maximum cc @dhiltgen @mxyng
GiteaMirror added the bug label 2026-05-04 20:46:46 -05:00
GiteaMirror commented 2026-05-04 20:46:56 -05:00
Author
Owner
Copy Link

@openzeka-birol-kuyumcu commented on GitHub (Sep 18, 2025):

for thor it works
Image

Image
<!-- gh-comment-id:3307559574 --> @openzeka-birol-kuyumcu commented on GitHub (Sep 18, 2025): for thor it works <img width="1920" height="1080" alt="Image" src="https://github.com/user-attachments/assets/f2774f8f-cf00-41b8-81e4-00aa4f354815" /> <img width="1920" height="1080" alt="Image" src="https://github.com/user-attachments/assets/0abab3ac-a489-49d4-95f3-6a8c8939463f" />
GiteaMirror commented 2026-05-04 20:46:58 -05:00
Author
Owner
Copy Link

@johnnynunez commented on GitHub (Sep 18, 2025):

for thor it works Image

Image

but with good performance?

<!-- gh-comment-id:3307770035 --> @johnnynunez commented on GitHub (Sep 18, 2025): > for thor it works <img alt="Image" width="1920" height="1080" src="https://private-user-images.githubusercontent.com/174419215/491108952-f2774f8f-cf00-41b8-81e4-00aa4f354815.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3NTgyMDU2NjIsIm5iZiI6MTc1ODIwNTM2MiwicGF0aCI6Ii8xNzQ0MTkyMTUvNDkxMTA4OTUyLWYyNzc0ZjhmLWNmMDAtNDFiOC04MWU0LTAwYWE0ZjM1NDgxNS5wbmc_WC1BbXotQWxnb3JpdGhtPUFXUzQtSE1BQy1TSEEyNTYmWC1BbXotQ3JlZGVudGlhbD1BS0lBVkNPRFlMU0E1M1BRSzRaQSUyRjIwMjUwOTE4JTJGdXMtZWFzdC0xJTJGczMlMkZhd3M0X3JlcXVlc3QmWC1BbXotRGF0ZT0yMDI1MDkxOFQxNDIyNDJaJlgtQW16LUV4cGlyZXM9MzAwJlgtQW16LVNpZ25hdHVyZT0wMTAzN2I0YjVmZDg3NTQxNzFmYjA2ZjZhNGNjNzE2OWU0Y2E0MjNkODU2YjFjYjU2NTYzNjg3ODVmMmFlNzYxJlgtQW16LVNpZ25lZEhlYWRlcnM9aG9zdCJ9.1_URsBU_T0_3coNZNe1ijpkYtmUsXxaCh8r0X0XJhNQ"> > > <img alt="Image" width="1920" height="1080" src="https://private-user-images.githubusercontent.com/174419215/491109017-0abab3ac-a489-49d4-95f3-6a8c8939463f.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3NTgyMDU2NjIsIm5iZiI6MTc1ODIwNTM2MiwicGF0aCI6Ii8xNzQ0MTkyMTUvNDkxMTA5MDE3LTBhYmFiM2FjLWE0ODktNDlkNC05NWYzLTZhOGM4OTM5NDYzZi5wbmc_WC1BbXotQWxnb3JpdGhtPUFXUzQtSE1BQy1TSEEyNTYmWC1BbXotQ3JlZGVudGlhbD1BS0lBVkNPRFlMU0E1M1BRSzRaQSUyRjIwMjUwOTE4JTJGdXMtZWFzdC0xJTJGczMlMkZhd3M0X3JlcXVlc3QmWC1BbXotRGF0ZT0yMDI1MDkxOFQxNDIyNDJaJlgtQW16LUV4cGlyZXM9MzAwJlgtQW16LVNpZ25hdHVyZT0wMmM5ZjRiMDkzZGM0OTdhNjUxMjEyYzA5NzdiM2I3ZjRjNGUxMjc5NzQwN2NiMDgwMjM0ZWQ0MDJmODQ3YzE5JlgtQW16LVNpZ25lZEhlYWRlcnM9aG9zdCJ9.gCxwKa2e7CKJWqvaFGGl9h2hNwj-uL8oJWMrKc071rs"> but with good performance?
GiteaMirror commented 2026-05-04 20:46:59 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Sep 18, 2025):

Can you share server logs?

<!-- gh-comment-id:3308292145 --> @dhiltgen commented on GitHub (Sep 18, 2025): Can you share server logs?
GiteaMirror commented 2026-05-04 20:47:00 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Sep 18, 2025):

I think I understand the problem. cuMemGetInfo_v2 is giving bad data on these systems and reports very little free memory available on the GPU after the system has been up for a while and the kernel is using system memory for buff/cache space. We wind up loading far fewer layers that could be loaded.

<!-- gh-comment-id:3308379329 --> @dhiltgen commented on GitHub (Sep 18, 2025): I think I understand the problem. cuMemGetInfo_v2 is giving bad data on these systems and reports very little free memory available on the GPU after the system has been up for a while and the kernel is using system memory for buff/cache space. We wind up loading far fewer layers that could be loaded.
GiteaMirror commented 2026-05-04 20:47:02 -05:00
Author
Owner
Copy Link

@openzeka-birol-kuyumcu commented on GitHub (Sep 19, 2025):

for thor it works Image
Image

but with good performance?

i think yes

Image

https://github.com/CordatusAI/ollama-benchmark

<!-- gh-comment-id:3310645534 --> @openzeka-birol-kuyumcu commented on GitHub (Sep 19, 2025): > > for thor it works <img alt="Image" width="1920" height="1080" src="https://private-user-images.githubusercontent.com/174419215/491108952-f2774f8f-cf00-41b8-81e4-00aa4f354815.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3NTgyMDU2NjIsIm5iZiI6MTc1ODIwNTM2MiwicGF0aCI6Ii8xNzQ0MTkyMTUvNDkxMTA4OTUyLWYyNzc0ZjhmLWNmMDAtNDFiOC04MWU0LTAwYWE0ZjM1NDgxNS5wbmc_WC1BbXotQWxnb3JpdGhtPUFXUzQtSE1BQy1TSEEyNTYmWC1BbXotQ3JlZGVudGlhbD1BS0lBVkNPRFlMU0E1M1BRSzRaQSUyRjIwMjUwOTE4JTJGdXMtZWFzdC0xJTJGczMlMkZhd3M0X3JlcXVlc3QmWC1BbXotRGF0ZT0yMDI1MDkxOFQxNDIyNDJaJlgtQW16LUV4cGlyZXM9MzAwJlgtQW16LVNpZ25hdHVyZT0wMTAzN2I0YjVmZDg3NTQxNzFmYjA2ZjZhNGNjNzE2OWU0Y2E0MjNkODU2YjFjYjU2NTYzNjg3ODVmMmFlNzYxJlgtQW16LVNpZ25lZEhlYWRlcnM9aG9zdCJ9.1_URsBU_T0_3coNZNe1ijpkYtmUsXxaCh8r0X0XJhNQ"> > > <img alt="Image" width="1920" height="1080" src="https://private-user-images.githubusercontent.com/174419215/491109017-0abab3ac-a489-49d4-95f3-6a8c8939463f.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3NTgyMDU2NjIsIm5iZiI6MTc1ODIwNTM2MiwicGF0aCI6Ii8xNzQ0MTkyMTUvNDkxMTA5MDE3LTBhYmFiM2FjLWE0ODktNDlkNC05NWYzLTZhOGM4OTM5NDYzZi5wbmc_WC1BbXotQWxnb3JpdGhtPUFXUzQtSE1BQy1TSEEyNTYmWC1BbXotQ3JlZGVudGlhbD1BS0lBVkNPRFlMU0E1M1BRSzRaQSUyRjIwMjUwOTE4JTJGdXMtZWFzdC0xJTJGczMlMkZhd3M0X3JlcXVlc3QmWC1BbXotRGF0ZT0yMDI1MDkxOFQxNDIyNDJaJlgtQW16LUV4cGlyZXM9MzAwJlgtQW16LVNpZ25hdHVyZT0wMmM5ZjRiMDkzZGM0OTdhNjUxMjEyYzA5NzdiM2I3ZjRjNGUxMjc5NzQwN2NiMDgwMjM0ZWQ0MDJmODQ3YzE5JlgtQW16LVNpZ25lZEhlYWRlcnM9aG9zdCJ9.gCxwKa2e7CKJWqvaFGGl9h2hNwj-uL8oJWMrKc071rs"> > > but with good performance? i think yes <img width="1411" height="710" alt="Image" src="https://github.com/user-attachments/assets/34a4cab6-ec57-460f-bd50-77f59dbc2bbb" /> https://github.com/CordatusAI/ollama-benchmark
GiteaMirror commented 2026-05-04 20:47:02 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Nov 6, 2025):

Memory reporting should be working properly now, so I think we can close this one.

<!-- gh-comment-id:3494200193 --> @dhiltgen commented on GitHub (Nov 6, 2025): Memory reporting should be working properly now, so I think we can close this one.
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#70249
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?