Skip to content
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

Is there a performance improvement to vectorized kernels when aligning memory? #7

Closed
MikaelSlevinsky opened this issue Jul 6, 2018 · 2 comments
Closed

Is there a performance improvement to vectorized kernels when aligning memory? #7

MikaelSlevinsky opened this issue Jul 6, 2018 · 2 comments
Assignees
@dawese
Labels
question Further information is requested

Comments

@MikaelSlevinsky
Copy link
Owner

No description provided.

@MikaelSlevinsky MikaelSlevinsky added the question Further information is requested label Jul 6, 2018
@dawese
Copy link
Contributor

dawese commented Jul 26, 2018

Due to the amount of conflicting information, I believe to best course of action is to implement an aligned version and compare performance results. Thus I shall pursue this COA and return with results.

@dawese
Copy link
Contributor

dawese commented Jul 27, 2018 •
edited
Loading

Preliminary testing suggests no. In fact aligned SSE seems to take and average of 3% longer to complete. I believe this suggests no appreciable difference between aligned vs unaligned data for our purposes.

Edit: Further testing reveals aligned AVX is about 14% faster!

@dawese dawese closed this as completed Jul 27, 2018
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@dawese dawese

Labels
question Further information is requested
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@MikaelSlevinsky @dawese