[GH-ISSUE #7457] Adding avx2+avx512 to cuda runner in new ollama code #51251

Closed
opened 2026-04-28 19:04:48 -05:00 by GiteaMirror · 0 comments
Owner

Originally created by @AncientMystic on GitHub (Oct 31, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/7457

Originally assigned to: @dhiltgen on GitHub.

In the old code i added avx2+512 in the gen_windows.ps1 by simply adding DGGML_AVX2=on & DGGML_AVX512=on after the DGGML_AVX=on line in the cuda build function

It added a fairly decent performance boost

I have added avx512 to cpu, But In the new code i cannot seem to find where to properly add it in the make files to also add it to cuda

What do i need to change on what lines in which files to add this to the new ollama code?

Originally created by @AncientMystic on GitHub (Oct 31, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/7457 Originally assigned to: @dhiltgen on GitHub. In the old code i added avx2+512 in the gen_windows.ps1 by simply adding DGGML_AVX2=on & DGGML_AVX512=on after the DGGML_AVX=on line in the cuda build function It added a fairly decent performance boost I have added avx512 to cpu, But In the new code i cannot seem to find where to properly add it in the make files to also add it to cuda What do i need to change on what lines in which files to add this to the new ollama code?
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#51251