__lasx_xvmulwod_q_du_d
LASX intrinsic · group: Integer Computation
Synopsis
__m256i __lasx_xvmulwod_q_du_d (__m256i a, __m256i b)
#include <lasxintrin.h>
Instruction: xvmulwod.q.du.d xr, xr, xr
CPU Flags: LASX
Description
Multiply the odd-positioned unsigned 64-bit elements in a and the corresponding signed 64-bit elements in b, extending both to 128-bit, and save the 128-bit result in dst.
Operation
for (int i = 0; i < 2; i++) {
dst.qword[i] = (u128)(u64)a.dword[2 * i + 1] * (s128)(s64)b.dword[2 * i + 1];
}
// Expands to:
if (0) {
dst.qword[0] = ((u128)((u64)a.dword[1])) * ((s128)((s64)b.dword[1]));
dst.qword[1] = ((u128)((u64)a.dword[3])) * ((s128)((s64)b.dword[3]));
}
Tested on real machine.
Latency and Throughput
| CPU | µarch | Latency | Throughput (IPC) |
|---|---|---|---|
| 3C5000 | LA464 | 7 | 2 |
| 3A6000 | LA664 | 7 | 2 |
| 3C6000 | LA664 | 7 | 2 |