__lsx_vpickev_h
LSX intrinsic · group: Misc
Synopsis
__m128i __lsx_vpickev_h (__m128i a, __m128i b)
#include <lsxintrin.h>
Instruction: vpickev.h vr, vr, vr
CPU Flags: LSX
Description
For each 128-bit lane, place even-positioned 16-bit elements of b into its lower half and even-positioned 16-bit elements of a into its upper half.
Operation
for (int i = 0; i < 8; i++) {
dst.half[i] = (i < 4) ? b.half[i * 2] : a.half[(i - 4) * 2];
}
// Expands to:
if (0) {
dst.half[0] = b.half[0];
dst.half[1] = b.half[2];
dst.half[2] = b.half[4];
dst.half[3] = b.half[6];
dst.half[4] = a.half[0];
dst.half[5] = a.half[2];
dst.half[6] = a.half[4];
dst.half[7] = a.half[6];
}
Tested on real machine.
Latency and Throughput
| CPU | µarch | Latency | Throughput (IPC) |
|---|---|---|---|
| 3C5000 | LA464 | 1 | 2 |
| 3A6000 | LA664 | 1 | 4 |
| 3C6000 | LA664 | 1 | 4 |
| 2K1000LA | LA264 | 1 | 1 |
| 2K3000 | LA364E | 1 | 1 |