github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-16 21:25:46 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[PR #6379] [CLOSED] Separate ARM64 CPU builds from x64 CPU builds and use Clang instead #22635

New Issue
Closed
opened 2026-04-19 16:27:07 -05:00 by GiteaMirror · 0 comments
GiteaMirror commented 2026-04-19 16:27:07 -05:00
Owner
Copy Link

📋 Pull Request Information

Original PR: https://github.com/ollama/ollama/pull/6379
Author: @hmartinez82
Created: 8/15/2024
Status: ❌ Closed

Base: main ← Head: arm64_armv87

📝 Commits (1)

  • 153ab64 Separate ARM64 CPU builds from x64 CPU builds and use Clang instead

📊 Changes

1 file changed (+74 additions, -4 deletions)

View changed files

📝 llm/generate/gen_windows.ps1 (+74 -4)

📄 Description

This enhances https://github.com/ollama/ollama/pull/3972 by making builds that are much, much faster:

  • Separate the ARM64 CPU builds from x64.
  • Make the Low Common Denominator CPU, for ARM64, to be ARMv8.2-A and use the matrix multiplication kernels tailored for NEON and ARMv8.2-A (See https://justine.lol/matmul/).
  • Create a separate runner built for ARMv8.7-A (e.g.: Snapdragon Plus/Elite X) that uses the MATMUL instructions.

The build instructions for ARM64 stay the same as in https://github.com/ollama/ollama/pull/5268. Clang was already required to build the llama.cpp static_library anyway so switching the runners to be built with Clang (on ARM64) doesn't add new requirements.

🔄 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/6379 **Author:** [@hmartinez82](https://github.com/hmartinez82) **Created:** 8/15/2024 **Status:** ❌ Closed **Base:** `main` ← **Head:** `arm64_armv87` --- ### 📝 Commits (1) - [`153ab64`](https://github.com/ollama/ollama/commit/153ab644672119e73529f31256b8e56c9ecd958c) Separate ARM64 CPU builds from x64 CPU builds and use Clang instead ### 📊 Changes **1 file changed** (+74 additions, -4 deletions) <details> <summary>View changed files</summary> 📝 `llm/generate/gen_windows.ps1` (+74 -4) </details> ### 📄 Description This enhances https://github.com/ollama/ollama/pull/3972 by making builds that are much, much faster: - Separate the ARM64 CPU builds from x64. - Make the Low Common Denominator CPU, for ARM64, to be ARMv8.2-A and use the matrix multiplication kernels tailored for NEON and ARMv8.2-A (See https://justine.lol/matmul/). - Create a separate runner built for ARMv8.7-A (e.g.: Snapdragon Plus/Elite X) that uses the MATMUL instructions. The build instructions for ARM64 stay the same as in https://github.com/ollama/ollama/pull/5268. Clang was already required to build the llama.cpp static_library anyway so switching the runners to be built with Clang (on ARM64) doesn't add new requirements. --- <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 2026-04-19 16:27:07 -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#22635
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?