Function vcvt_high_f16_f32

#[target_feature(enable = "neon")]
pub fn vcvt_high_f16_f32(a: float16x4_t, b: float32x4_t) -> float16x8_t

Floating-point convert to lower precision Arm's documentation