-
Notifications
You must be signed in to change notification settings - Fork 8
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
Use SIMDe #22
Conversation
1b09b1e
to
29262a3
Compare
29262a3
to
9b86198
Compare
450e585
to
704a52f
Compare
- created build.hpp for TPUBuild - epu8id and similar are now Epu8.id() - improved doc
check_cxx_compiler_flag('-mavx' HPCOMBI_HAVE_FLAG_AVX) | ||
#check_cxx_compiler_flag('-march=native' HPCOMBI_HAVE_FLAG_NATIVE) | ||
#check_cxx_compiler_flag('-mavx2' HPCOMBI_HAVE_FLAG_AVX2) | ||
#check_cxx_compiler_flag('-mavx512bw' HPCOMBI_HAVE_FLAG_AVX512BW) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I recommend also checking for -fopenmp-simd
and if it is found, adding it and -DSIMDE_ENABLE_OPENMP
(generates better autovectorization)
README.md
Outdated
HPCombi was initially designed using the SSE and AVX instruction sets, and did | ||
not work on machines without these instructions (such as ARM). From v1.0.0 | ||
HPCombi supports processors with other instruction sets also, via | ||
[simd-everywhere](https://github.com/simd-everywhere/simde). It might be the | ||
case that the greatest performance gains are achieved on processors supporting | ||
the SSE and AVX instruction sets, but the HPCombi benchmarks indicate that | ||
there are also still signficant gains on other processors too. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I usually say SIMDe
or SIMD Everywhere (SIMDe)
.
This is WIP but I wanted to see if the tests can be run from this branch in the CI.