pub unsafe fn _mm256_moveldup_ps(a: __m256) -> __m256
Duplicate even-indexed single-precision (32-bit) floating-point elements from a, and returns the results.
a
Intel’s documentation