github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-08 17:49:24 -05:00
Code Issues 14.6k Packages Projects Releases 122 Wiki Activity

[PR #4420] [CLOSED] use nvml for gpu info #37358

New Issue
Closed
opened 2026-04-22 22:04:27 -05:00 by GiteaMirror · 0 comments
GiteaMirror commented 2026-04-22 22:04:27 -05:00
Owner
Copy Link

📋 Pull Request Information

Original PR: https://github.com/ollama/ollama/pull/4420
Author: @jmorganca
Created: 5/14/2024
Status: ❌ Closed

Base: main ← Head: jmorganca/nvml

📝 Commits (1)

📊 Changes

7 files changed (+279 additions, -16 deletions)

View changed files

📝 gpu/gpu.go (+62 -8)
📝 gpu/gpu_info.h (+1 -0)
📝 gpu/gpu_info_nvcuda.c (+28 -7)
📝 gpu/gpu_info_nvcuda.h (+4 -0)
➕ gpu/gpu_info_nvml.c (+112 -0)
➕ gpu/gpu_info_nvml.h (+71 -0)
📝 server/sched.go (+1 -1)

📄 Description

Also decreases period between recovery lookups since memory is freed quickly on graphics cards like the 4090

TODO:

  • Test on Windows 10
  • Test on other Nvidia GPUs
  • Test on Linux
  • Test on WSL2
  • Make sure nvml.dll (and equivalent .so) lookup paths are correct (it seems that it will always be in C:\Windows\System32 on Windows)
  • rocsmi for AMD?

🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.

## 📋 Pull Request Information **Original PR:** https://github.com/ollama/ollama/pull/4420 **Author:** [@jmorganca](https://github.com/jmorganca) **Created:** 5/14/2024 **Status:** ❌ Closed **Base:** `main` ← **Head:** `jmorganca/nvml` --- ### 📝 Commits (1) - [`9136080`](https://github.com/ollama/ollama/commit/9136080aeb7fad71ee00ee724117615c5ee62aac) use nvml for gpu info ### 📊 Changes **7 files changed** (+279 additions, -16 deletions) <details> <summary>View changed files</summary> 📝 `gpu/gpu.go` (+62 -8) 📝 `gpu/gpu_info.h` (+1 -0) 📝 `gpu/gpu_info_nvcuda.c` (+28 -7) 📝 `gpu/gpu_info_nvcuda.h` (+4 -0) ➕ `gpu/gpu_info_nvml.c` (+112 -0) ➕ `gpu/gpu_info_nvml.h` (+71 -0) 📝 `server/sched.go` (+1 -1) </details> ### 📄 Description Also decreases period between recovery lookups since memory is freed quickly on graphics cards like the 4090 TODO: - [ ] Test on Windows 10 - [ ] Test on other Nvidia GPUs - [ ] Test on Linux - [ ] Test on WSL2 - [ ] Make sure `nvml.dll` (and equivalent .so) lookup paths are correct (it seems that it will always be in `C:\Windows\System32` on Windows) - [ ] rocsmi for AMD? --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>
GiteaMirror added the pull-request label 2026-04-22 22:04:27 -05:00
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 pull-request
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#37358
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?