KMOVW (K, K) - Latency

Operands

Operand 1 (w): Register (K0, K1, K2, K3, K4, K5, K6, K7)
Operand 2 (r): Register (K0, K1, K2, K3, K4, K5, K6, K7)

Latency operand 2 → 1: 1

Latency operand 2 → 1: 1

Experiment 1

Instruction: KMOVW K2, K1
Chain instruction: KMOVW K1, K2

Code:

   0:	c5 f8 90 d1          	kmovw  k2,k1
   4:	c5 f8 90 ca          	kmovw  k1,k2

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL

Show nanoBench command
Results:
- Instructions retired: 2.0
- Core cycles: 2.0
- Reference cycles: 1.1
- UOPS_EXECUTED.THREAD: 2.0

Experiment 2

Instruction: KMOVW K2, K1
Chain instructions: KMOVQ K1, K2;KMOVQ K1, K1;KMOVQ K1, K1;KMOVQ K1, K1;KMOVQ K1, K1
Chain latency: 5

Code:

   0:	c5 f8 90 d1          	kmovw  k2,k1
   4:	c4 e1 f8 90 ca       	kmovq  k1,k2
   9:	c4 e1 f8 90 c9       	kmovq  k1,k1
   e:	c4 e1 f8 90 c9       	kmovq  k1,k1
  13:	c4 e1 f8 90 c9       	kmovq  k1,k1
  18:	c4 e1 f8 90 c9       	kmovq  k1,k1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL

Show nanoBench command
Results:
- Instructions retired: 6.0
- Core cycles: 6.0
- Reference cycles: 3.41
- UOPS_EXECUTED.THREAD: 6.0