Vectorization of real numbers using AVX with blocks of 8 reals.
More...
#include <arcane/utils/SimdAVX.h>
Vectorization of real numbers using AVX with blocks of 8 reals.
- Note
- This class must be aligned to 32 bytes.
Definition at line 245 of file SimdAVX.h.
◆ anonymous enum
◆ AVXSimdX8Real() [1/6]
| Arcane::AVXSimdX8Real::AVXSimdX8Real |
( |
| ) |
|
|
inline |
◆ AVXSimdX8Real() [2/6]
| Arcane::AVXSimdX8Real::AVXSimdX8Real |
( |
__m256d | _v0, |
|
|
__m256d | _v1 ) |
|
inline |
◆ AVXSimdX8Real() [3/6]
| Arcane::AVXSimdX8Real::AVXSimdX8Real |
( |
Real | r | ) |
|
|
inlineexplicit |
◆ AVXSimdX8Real() [4/6]
◆ AVXSimdX8Real() [5/6]
| Arcane::AVXSimdX8Real::AVXSimdX8Real |
( |
const Real * | base, |
|
|
const Int32 * | idx ) |
|
inline |
◆ AVXSimdX8Real() [6/6]
| Arcane::AVXSimdX8Real::AVXSimdX8Real |
( |
const Real * | base | ) |
|
|
inlineexplicit |
Loads contiguous values located at address base, which must be aligned.
Definition at line 297 of file SimdAVX.h.
◆ fromScalar()
◆ operator-()
| AVXSimdX8Real Arcane::AVXSimdX8Real::operator- |
( |
| ) |
const |
|
inline |
◆ operator[]() [1/2]
◆ operator[]() [2/2]
| Real Arcane::AVXSimdX8Real::operator[] |
( |
Integer | i | ) |
const |
|
inline |
◆ set() [1/2]
| void Arcane::AVXSimdX8Real::set |
( |
ARCANE_RESTRICT Real * | base | ) |
const |
|
inline |
Stores the instance values at address base, which must be aligned.
Definition at line 329 of file SimdAVX.h.
◆ set() [2/2]
| void Arcane::AVXSimdX8Real::set |
( |
ARCANE_RESTRICT Real * | base, |
|
|
const ARCANE_RESTRICT Int32 * | idx ) const |
|
inline |
◆ BLOCK_SIZE
| const int Arcane::AVXSimdX8Real::BLOCK_SIZE = 8 |
|
static |
◆ v0
| __m256d Arcane::AVXSimdX8Real::v0 |
◆ v1
| __m256d Arcane::AVXSimdX8Real::v1 |
The documentation for this class was generated from the following file: