Make CAReduce more SIMD and memory friendly #1385

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Draft

ricardoV94 wants to merge 1 commit into pymc-devs:main from ricardoV94:faster_careduce

Member

ricardoV94 commented Apr 30, 2025 •

edited by github-actions bot

Loading

This PR does two major changes:

Adds a branch with intermediate allocators when reducing over a contiguous dimension that is SIMD friendly.
Allocate output buffer aligned with input dimensions, instead of always allocating C-order.

Performance is now comparable or better than numpy

📚 Documentation preview 📚: https://pytensor--1385.org.readthedocs.build/en/1385/


          Make CAReduce more SIMD friendly and do better allocation of output m…

3ef6fb8

…emory

ricardoV94 force-pushed the faster_careduce branch from 293af91 to 3ef6fb8 Compare

May 1, 2025 19:44

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet