VMOVSD (XMM, M64) - Latency

Experiment 1

• Instruction: VMOVSD XMM0, qword ptr [R14]
• Chain instructions: VMOVQ R12, XMM0;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12
• Chain latency: ≥5
• Code:

     0:	c4 c1 7b 10 06       	vmovsd xmm0,QWORD PTR [r14]
     5:	c4 c1 f9 7e c4       	vmovq  r12,xmm0
     a:	4d 31 e6             	xor    r14,r12
     d:	4d 31 e6             	xor    r14,r12
    10:	4d 31 e6             	xor    r14,r12
    13:	4d 31 e6             	xor    r14,r12
  

• Init:

  MOV R15, 10000;
  L: VADDPS YMM0, YMM1, YMM1;
  VADDPS YMM0, YMM1, YMM1;
  DEC R15;
  JNZ L;
  VZEROALL;
  MOV RAX, 0x4000000040000000;
  MOV [R14+0], RAX

• Show nanoBench command
• Results:
  • Instructions retired: 6.0
  • Core cycles: 11.0
  • Reference cycles: 10.93
  • UOPS_EXECUTED.CORE: 6.0

Experiment 1

• Instruction: VMOVSD XMM0, qword ptr [R14+R13]
• Chain instructions: VMOVQ R12, XMM0;XOR R13, R12;XOR R13, R12;XOR R13, R12;XOR R13, R12
• Chain latency: ≥5
• Code:

     0:	c4 81 7b 10 04 2e    	vmovsd xmm0,QWORD PTR [r14+r13*1]
     6:	c4 c1 f9 7e c4       	vmovq  r12,xmm0
     b:	4d 31 e5             	xor    r13,r12
     e:	4d 31 e5             	xor    r13,r12
    11:	4d 31 e5             	xor    r13,r12
    14:	4d 31 e5             	xor    r13,r12
  

• Init:

  MOV R15, 10000;
  L: VADDPS YMM0, YMM1, YMM1;
  VADDPS YMM0, YMM1, YMM1;
  DEC R15;
  JNZ L;
  VZEROALL;
  MOV RAX, 0x4000000040000000;
  MOV [R14+0], RAX;
  XOR R13, R13

• Show nanoBench command
• Results:
  • Instructions retired: 6.0
  • Core cycles: 11.0
  • Reference cycles: 10.99
  • UOPS_EXECUTED.CORE: 6.0

Experiment 1

• Instruction: VMOVSD XMM0, qword ptr [R14]
• Chain instructions: VSHUFPD XMM13, XMM0, XMM0, 0;VSHUFPD XMM13, XMM13, XMM13, 0;VSHUFPD XMM13, XMM13, XMM13, 0;VSHUFPD XMM13, XMM13, XMM13, 0;VSHUFPD XMM13, XMM13, XMM13, 0;VMOVSD [R14], XMM13
• Chain latency: ≥6
• Code:

     0:	c4 c1 7b 10 06       	vmovsd xmm0,QWORD PTR [r14]
     5:	c5 79 c6 e8 00       	vshufpd xmm13,xmm0,xmm0,0x0
     a:	c4 41 11 c6 ed 00    	vshufpd xmm13,xmm13,xmm13,0x0
    10:	c4 41 11 c6 ed 00    	vshufpd xmm13,xmm13,xmm13,0x0
    16:	c4 41 11 c6 ed 00    	vshufpd xmm13,xmm13,xmm13,0x0
    1c:	c4 41 11 c6 ed 00    	vshufpd xmm13,xmm13,xmm13,0x0
    22:	c4 41 7b 11 2e       	vmovsd QWORD PTR [r14],xmm13
  

• Init:

  MOV R15, 10000;
  L: VADDPS YMM0, YMM1, YMM1;
  VADDPS YMM0, YMM1, YMM1;
  DEC R15;
  JNZ L;
  VZEROALL;
  MOV RAX, 0x4000000040000000;
  MOV [R14+0], RAX

• Show nanoBench command
• Results:
  • Instructions retired: 7.0
  • Core cycles: 11.0
  • Reference cycles: 10.93
  • UOPS_EXECUTED.CORE: 8.0

Experiment 2

• Instruction: VMOVSD XMM0, qword ptr [R14]
• Chain instructions: VPAND XMM13, XMM0, XMM0;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VMOVQ [R14], XMM13
• Chain latency: ≥6
• Code:

     0:	c4 c1 7b 10 06       	vmovsd xmm0,QWORD PTR [r14]
     5:	c5 79 db e8          	vpand  xmm13,xmm0,xmm0
     9:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
     e:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
    13:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
    18:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
    1d:	c4 41 79 d6 2e       	vmovq  QWORD PTR [r14],xmm13
  

• Init:

  MOV R15, 10000;
  L: VADDPS YMM0, YMM1, YMM1;
  VADDPS YMM0, YMM1, YMM1;
  DEC R15;
  JNZ L;
  VZEROALL;
  MOV RAX, 0x4000000040000000;
  MOV [R14+0], RAX

• Show nanoBench command
• Results:
  • Instructions retired: 7.0
  • Core cycles: 11.0
  • Reference cycles: 11.04
  • UOPS_EXECUTED.CORE: 8.0