github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-17 13:47:08 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[PR #5570] [CLOSED] New intel gpus detect backend #17101

New Issue
Closed
opened 2026-04-16 05:52:45 -05:00 by GiteaMirror · 0 comments
GiteaMirror commented 2026-04-16 05:52:45 -05:00
Owner
Copy Link

📋 Pull Request Information

Original PR: https://github.com/ollama/ollama/pull/5570
Author: @zhewang1-intc
Created: 7/9/2024
Status: ❌ Closed

Base: main ← Head: new_intel_gpus_detect_backend

📝 Commits (3)

📊 Changes

9 files changed (+173 additions, -455 deletions)

View changed files

📝 gpu/gpu.go (+32 -40)
📝 gpu/gpu_info_oneapi.c (+13 -224)
📝 gpu/gpu_info_oneapi.h (+28 -187)
📝 gpu/gpu_linux.go (+3 -2)
➕ gpu/libollama_detect_intel_gpu.so (+0 -0)
📝 gpu/types.go (+1 -2)
➕ llm/generate/.clang-format (+7 -0)
➕ llm/generate/sycl_intel_gpu_mgr.cpp (+43 -0)
➕ llm/generate/sycl_intel_gpu_mgr.hpp (+46 -0)

📄 Description

Hi @dhiltgen , I am trying to make Intel GPU more productive on Ollama recently. This PR is still under development, and the overall structure may undergo significant changes in the future.

This PR mainly includes two points:

  1. Support Intel iGPUs. Intel took the lead in introducing the concept of AIPC and launched the Core Ultra processor with a powerful iGPU. We believe that using the iGPU for inference on the Core Ultra client CPU may bring better performance in the future.
  2. Refactoring the detection of Intel GPU. Switching from the level-zero API to the SYCL API, there will be a 1D device list query by sycl api rather a 2D matrix like the level-zero API. This will provide richer device information and a more convenient detection process.

🔄 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/5570 **Author:** [@zhewang1-intc](https://github.com/zhewang1-intc) **Created:** 7/9/2024 **Status:** ❌ Closed **Base:** `main` ← **Head:** `new_intel_gpus_detect_backend` --- ### 📝 Commits (3) - [`3a0a90b`](https://github.com/ollama/ollama/commit/3a0a90bb2e869996eaa942006a248a8bc3397dad) new backend draft - [`622fb98`](https://github.com/ollama/ollama/commit/622fb981cecd396aaf9772783e16871342451e5b) draft works on arc770m - [`b87d0e4`](https://github.com/ollama/ollama/commit/b87d0e4ad75f29edc65111ec6dafbd477060ecfb) fix warning & better naming ### 📊 Changes **9 files changed** (+173 additions, -455 deletions) <details> <summary>View changed files</summary> 📝 `gpu/gpu.go` (+32 -40) 📝 `gpu/gpu_info_oneapi.c` (+13 -224) 📝 `gpu/gpu_info_oneapi.h` (+28 -187) 📝 `gpu/gpu_linux.go` (+3 -2) ➕ `gpu/libollama_detect_intel_gpu.so` (+0 -0) 📝 `gpu/types.go` (+1 -2) ➕ `llm/generate/.clang-format` (+7 -0) ➕ `llm/generate/sycl_intel_gpu_mgr.cpp` (+43 -0) ➕ `llm/generate/sycl_intel_gpu_mgr.hpp` (+46 -0) </details> ### 📄 Description Hi @dhiltgen , I am trying to make Intel GPU more productive on Ollama recently. This PR is still under development, and the overall structure may undergo significant changes in the future. This PR mainly includes two points: 1. Support Intel iGPUs. Intel took the lead in introducing the concept of AIPC and launched the Core Ultra processor with a powerful iGPU. We believe that using the iGPU for inference on the Core Ultra client CPU may bring better performance in the future. 2. Refactoring the detection of Intel GPU. Switching from the level-zero API to the SYCL API, there will be a 1D device list query by sycl api rather a 2D matrix like the level-zero API. This will provide richer device information and a more convenient detection process. --- <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 2026-04-16 05:52:45 -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#17101
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?