__lasx_xvdiv_d
LASX intrinsic · group: Integer Computation
Synopsis
__m256i __lasx_xvdiv_d (__m256i a, __m256i b)
#include <lasxintrin.h>
Instruction: xvdiv.d xr, xr, xr
CPU Flags: LASX
Description
Divide each signed 64-bit element in a by the corresponding element in b.
Operation
for (int i = 0; i < 4; i++) {
dst.dword[i] = (b.dword[i] == 0) ? 0 : ((s64)a.dword[i] / (s64)b.dword[i]);
}
Tested on real machine.
Latency and Throughput
| CPU | µarch | Latency | Throughput (IPC) |
|---|---|---|---|
| 3C5000 | LA464 | 8, 18.5 | 0.11(1/9) |
| 3A6000 | LA664 | 8 | 0.25(1/4) |
| 3C6000 | LA664 | 18.5, 19 | 0.12(1/8.5) |