Skip to content

SIMD Vectorization with OpenMP

Jump to bottom
Man-Long Wong edited this page Apr 18, 2018 · 5 revisions

The code is fully vectorized with the OpenMP SIMD construct. To turn on the SIMD vectorization for speed-up, you will need to add the flag -DHAMERS_ENABLE_SIMD=ON when the cmake command is executed. The SIMD vectorization is turned off by default. The SIMD construct is also ignored in debug mode since OpenMP is not used.

Clone this wiki locally