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.

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

Implement some vec instructions with NEON #1122

Open
wants to merge 1 commit into
base: master
Choose a base branch
from

Conversation

nimelehin
Copy link
Contributor

@nimelehin nimelehin commented Dec 13, 2020

Test results (less is faster):

 CPU vec_add_b128(ms) vec_add_b128_neon(ms) diff vec_min_ub128(ms) vec_min_ub128_neon(ms) diff vec_compare_eqb128(ms) vec_compare_eqb128_neon(ms) diff vec_movmask_b128 vec_movmask_b128_neon diff
A14 40096,8 36707,0 +9,2% 63393,6 36696,8 +72,7% 40037,8 36717,0 +9,04% 28386,0 10013,0 +183,49%
A13 60300,0 41331,4 +45,9% 71451,8 41417,6 +72,5% 60226,2 41406,8 +45,5% 32007,0 11307,0 +183,07%
A12x 65126,2 44755,6 +45,5% 78699,4 45598,2 +72,6% 66397,8 45160,4 +47,0% 34102,0 12025,0 +183,59%

@tbodt
Copy link
Member

tbodt commented Dec 13, 2020

What effect does this have on non-micro benchmarks?

@nimelehin
Copy link
Contributor Author

nimelehin commented Dec 15, 2020

What effect does this have on non-micro benchmarks?

As a test I ran a simple C program working with sse, the performance is about 10-15%, but that seems to be a micro-bench too.
Tried to run GO, it uses sse actively, but no performance difference is seen in go1 test suite, according to profiler sse emulation doesn't take a lot of time comparing to other parts of emulation.

emkey1 pushed a commit to emkey1/ish-AOK that referenced this pull request Dec 27, 2021
… AOK

Added nimelehin's SSE enhancements from pending mainline/upstream PR ish-app#1122
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants