pub unsafe fn _mm256_mask_cmpneq_epu32_mask(
    k1: __mmask8,
    a: __m256i,
    b: __m256i
) -> __mmask8
🔬 This is a nightly-only experimental API. (stdsimd #48556)
Available on x86-64 and target feature avx512f,avx512vl only.
Expand description

Compare packed unsigned 32-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set).

Intel’s documentation