VMOVD (R32, XMM) - Latency

Experiment 1

• Instruction: VMOVD R8D, XMM0
• Chain instruction: VCVTSI2SD XMM0, XMM1, R8D
• Chain latency: ≥1
• Code:

     0:	c4 c1 79 7e c0       	vmovd  r8d,xmm0
     5:	c4 c1 73 2a c0       	vcvtsi2sd xmm0,xmm1,r8d
  

• Init:

  MOV R15, 10000;
  L: VADDPS YMM0, YMM1, YMM1;
  VADDPS YMM0, YMM1, YMM1;
  DEC R15;
  JNZ L;
  VZEROALL;
  VXORPS XMM0, XMM0, XMM0

• Show nanoBench command
• Results:
  • MPERF: 8.4
  • APERF: 10.0
  • UOPS: 3.0

Experiment 2

• Instruction: VMOVD R8D, XMM0
• Chain instruction: VCVTSI2SD XMM0, XMM1, R8
• Chain latency: ≥1
• Code:

     0:	c4 c1 79 7e c0       	vmovd  r8d,xmm0
     5:	c4 c1 f3 2a c0       	vcvtsi2sd xmm0,xmm1,r8
  

• Init:

  MOV R15, 10000;
  L: VADDPS YMM0, YMM1, YMM1;
  VADDPS YMM0, YMM1, YMM1;
  DEC R15;
  JNZ L;
  VZEROALL;
  VXORPS XMM0, XMM0, XMM0

• Show nanoBench command
• Results:
  • MPERF: 8.46
  • APERF: 10.0
  • UOPS: 3.0

Experiment 3

• Instruction: VMOVD R8D, XMM0
• Chain instruction: VCVTSI2SS XMM0, XMM1, R8D
• Chain latency: ≥1
• Code:

     0:	c4 c1 79 7e c0       	vmovd  r8d,xmm0
     5:	c4 c1 72 2a c0       	vcvtsi2ss xmm0,xmm1,r8d
  

• Init:

  MOV R15, 10000;
  L: VADDPS YMM0, YMM1, YMM1;
  VADDPS YMM0, YMM1, YMM1;
  DEC R15;
  JNZ L;
  VZEROALL;
  VXORPS XMM0, XMM0, XMM0

• Show nanoBench command
• Results:
  • MPERF: 8.4
  • APERF: 10.0
  • UOPS: 3.0

Experiment 4

• Instruction: VMOVD R8D, XMM0
• Chain instruction: VCVTSI2SS XMM0, XMM1, R8
• Chain latency: ≥1
• Code:

     0:	c4 c1 79 7e c0       	vmovd  r8d,xmm0
     5:	c4 c1 f2 2a c0       	vcvtsi2ss xmm0,xmm1,r8
  

• Init:

  MOV R15, 10000;
  L: VADDPS YMM0, YMM1, YMM1;
  VADDPS YMM0, YMM1, YMM1;
  DEC R15;
  JNZ L;
  VZEROALL;
  VXORPS XMM0, XMM0, XMM0

• Show nanoBench command
• Results:
  • MPERF: 8.4
  • APERF: 10.0
  • UOPS: 3.0

Experiment 5

• Instruction: VMOVD R8D, XMM0
• Chain instruction: VMOVD XMM0, R8D
• Chain latency: ≥1
• Code:

     0:	c4 c1 79 7e c0       	vmovd  r8d,xmm0
     5:	c4 c1 79 6e c0       	vmovd  xmm0,r8d
  

• Init:

  MOV R15, 10000;
  L: VADDPS YMM0, YMM1, YMM1;
  VADDPS YMM0, YMM1, YMM1;
  DEC R15;
  JNZ L;
  VZEROALL;
  VXORPS XMM0, XMM0, XMM0

• Show nanoBench command
• Results:
  • MPERF: 5.04
  • APERF: 6.0
  • UOPS: 2.0

Experiment 6

• Instruction: VMOVD R8D, XMM0
• Chain instruction: VMOVQ XMM0, R8
• Chain latency: ≥1
• Code:

     0:	c4 c1 79 7e c0       	vmovd  r8d,xmm0
     5:	c4 c1 f9 6e c0       	vmovq  xmm0,r8
  

• Init:

  MOV R15, 10000;
  L: VADDPS YMM0, YMM1, YMM1;
  VADDPS YMM0, YMM1, YMM1;
  DEC R15;
  JNZ L;
  VZEROALL;
  VXORPS XMM0, XMM0, XMM0

• Show nanoBench command
• Results:
  • MPERF: 5.04
  • APERF: 6.0
  • UOPS: 2.0

Experiment 7

• Instruction: VMOVD R8D, XMM0
• Chain instruction: VPINSRB XMM0, XMM1, R8D, 2
• Chain latency: ≥1
• Code:

     0:	c4 c1 79 7e c0       	vmovd  r8d,xmm0
     5:	c4 c3 71 20 c0 02    	vpinsrb xmm0,xmm1,r8d,0x2
  

• Init:

  MOV R15, 10000;
  L: VADDPS YMM0, YMM1, YMM1;
  VADDPS YMM0, YMM1, YMM1;
  DEC R15;
  JNZ L;
  VZEROALL;
  VXORPS XMM0, XMM0, XMM0

• Show nanoBench command
• Results:
  • MPERF: 5.94
  • APERF: 6.99
  • UOPS: 3.0

Experiment 8

• Instruction: VMOVD R8D, XMM0
• Chain instruction: VPINSRD XMM0, XMM1, R8D, 2
• Chain latency: ≥1
• Code:

     0:	c4 c1 79 7e c0       	vmovd  r8d,xmm0
     5:	c4 c3 71 22 c0 02    	vpinsrd xmm0,xmm1,r8d,0x2
  

• Init:

  MOV R15, 10000;
  L: VADDPS YMM0, YMM1, YMM1;
  VADDPS YMM0, YMM1, YMM1;
  DEC R15;
  JNZ L;
  VZEROALL;
  VXORPS XMM0, XMM0, XMM0

• Show nanoBench command
• Results:
  • MPERF: 6.0
  • APERF: 6.99
  • UOPS: 3.0

Experiment 9

• Instruction: VMOVD R8D, XMM0
• Chain instruction: VPINSRQ XMM0, XMM1, R8, 2
• Chain latency: ≥1
• Code:

     0:	c4 c1 79 7e c0       	vmovd  r8d,xmm0
     5:	c4 c3 f1 22 c0 02    	vpinsrq xmm0,xmm1,r8,0x2
  

• Init:

  MOV R15, 10000;
  L: VADDPS YMM0, YMM1, YMM1;
  VADDPS YMM0, YMM1, YMM1;
  DEC R15;
  JNZ L;
  VZEROALL;
  VXORPS XMM0, XMM0, XMM0

• Show nanoBench command
• Results:
  • MPERF: 7.2
  • APERF: 6.99
  • UOPS: 3.0

Experiment 10

• Instruction: VMOVD R8D, XMM0
• Chain instruction: VPINSRW XMM0, XMM1, R8D, 2
• Chain latency: ≥1
• Code:

     0:	c4 c1 79 7e c0       	vmovd  r8d,xmm0
     5:	c4 c1 71 c4 c0 02    	vpinsrw xmm0,xmm1,r8d,0x2
  

• Init:

  MOV R15, 10000;
  L: VADDPS YMM0, YMM1, YMM1;
  VADDPS YMM0, YMM1, YMM1;
  DEC R15;
  JNZ L;
  VZEROALL;
  VXORPS XMM0, XMM0, XMM0

• Show nanoBench command
• Results:
  • MPERF: 5.82
  • APERF: 6.99
  • UOPS: 3.0