__lasx_xvsubwev_w_hu

LASX intrinsic · group: Integer Computation

Synopsis

__m256i __lasx_xvsubwev_w_hu (__m256i a, __m256i b)
#include <lasxintrin.h>
Instruction: xvsubwev.w.hu xr, xr, xr
CPU Flags: LASX

Description

Subtract the even-positioned unsigned 16-bit elements in b from the corresponding elements in a, extending both to 32-bit, and save the 32-bit result in dst.

Operation

for (int i = 0; i < 8; i++) {
  dst.word[i] = (u32)(u16)a.half[2 * i] - (u32)(u16)b.half[2 * i];
}

// Expands to:

if (0) {
  dst.word[0] = ((u32)((u16)a.half[0])) - ((u32)((u16)b.half[0]));
  dst.word[1] = ((u32)((u16)a.half[2])) - ((u32)((u16)b.half[2]));
  dst.word[2] = ((u32)((u16)a.half[4])) - ((u32)((u16)b.half[4]));
  dst.word[3] = ((u32)((u16)a.half[6])) - ((u32)((u16)b.half[6]));
  dst.word[4] = ((u32)((u16)a.half[8])) - ((u32)((u16)b.half[8]));
  dst.word[5] = ((u32)((u16)a.half[10])) - ((u32)((u16)b.half[10]));
  dst.word[6] = ((u32)((u16)a.half[12])) - ((u32)((u16)b.half[12]));
  dst.word[7] = ((u32)((u16)a.half[14])) - ((u32)((u16)b.half[14]));
}

Tested on real machine.

Latency and Throughput

CPU	µarch	Latency	Throughput (IPC)
3C5000	LA464	2	2
3A6000	LA664	2	2
3C6000	LA664	2	2

---

All intrinsics in alphabetical order