pub unsafe fn _mm256_cvtepi32_ps(a: __m256i) -> __m256
Converts packed 32-bit integers in a to packed single-precision (32-bit) floating-point elements.
a
Intel’s documentation