__lasx_xvmulwev_w_hu_h
LASX intrinsic · group: Integer Computation
Synopsis
__m256i __lasx_xvmulwev_w_hu_h (__m256i a, __m256i b)
#include <lasxintrin.h>
Instruction: xvmulwev.w.hu.h xr, xr, xr
CPU Flags: LASX
Description
Multiply the even-positioned unsigned 16-bit elements in a and the corresponding signed 16-bit elements in b, 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] * (s32)(s16)b.half[2 * i];
}
// Expands to:
if (0) {
dst.word[0] = ((u32)((u16)a.half[0])) * ((s32)((s16)b.half[0]));
dst.word[1] = ((u32)((u16)a.half[2])) * ((s32)((s16)b.half[2]));
dst.word[2] = ((u32)((u16)a.half[4])) * ((s32)((s16)b.half[4]));
dst.word[3] = ((u32)((u16)a.half[6])) * ((s32)((s16)b.half[6]));
dst.word[4] = ((u32)((u16)a.half[8])) * ((s32)((s16)b.half[8]));
dst.word[5] = ((u32)((u16)a.half[10])) * ((s32)((s16)b.half[10]));
dst.word[6] = ((u32)((u16)a.half[12])) * ((s32)((s16)b.half[12]));
dst.word[7] = ((u32)((u16)a.half[14])) * ((s32)((s16)b.half[14]));
}
Tested on real machine.
Latency and Throughput
| CPU | µarch | Latency | Throughput (IPC) |
|---|---|---|---|
| 3C5000 | LA464 | 4 | 2 |
| 3A6000 | LA664 | 4 | 2 |
| 3C6000 | LA664 | 4 | 2 |