pub fn _mm_cvt_roundss_sh(a: __m128h, b: __m128, const ROUNDING: i32) -> __m128h🔬This is a nightly-only experimental API. (
stdarch_x86_avx512_f16 #127213)Available on (x86 or x86-64) and target feature
avx512fp16 and x86-64 only.Expand description
Convert the lower single-precision (32-bit) floating-point element in b to a half-precision (16-bit) floating-point elements, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst.
Rounding is done according to the rounding parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT|_MM_FROUND_NO_EXC: round to nearest and suppress exceptions_MM_FROUND_TO_NEG_INF|_MM_FROUND_NO_EXC: round down and suppress exceptions_MM_FROUND_TO_POS_INF|_MM_FROUND_NO_EXC: round up and suppress exceptions_MM_FROUND_TO_ZERO|_MM_FROUND_NO_EXC: truncate and suppress exceptions_MM_FROUND_CUR_DIRECTION: useMXCSR.RC- see_MM_SET_ROUNDING_MODE