MOVZX (R16, M8) - Latency

Experiment 1

• Instruction: MOVZX R8W, byte ptr [R14]
• Code:

     0:	66 45 0f b6 06       	movzx  r8w,BYTE PTR [r14]
  

• Show nanoBench command
• Results:
  • Instructions retired: 1.0
  • Core cycles: 1.0
  • Reference cycles: 0.96
  • UOPS_EXECUTED.CORE: 2.0

Experiment 2

• Instruction: MOVZX R8W, byte ptr [R14]
• Chain instructions: MOVSX R8, R8W;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D
• Chain latency: 5
• Code:

     0:	66 45 0f b6 06       	movzx  r8w,BYTE PTR [r14]
     5:	4d 0f bf c0          	movsx  r8,r8w
     9:	4d 63 c0             	movsxd r8,r8d
     c:	4d 63 c0             	movsxd r8,r8d
     f:	4d 63 c0             	movsxd r8,r8d
    12:	4d 63 c0             	movsxd r8,r8d
  

• Show nanoBench command
• Results:
  • Instructions retired: 6.0
  • Core cycles: 6.0
  • Reference cycles: 5.39
  • UOPS_EXECUTED.CORE: 7.0

Experiment 1

• Instruction: MOVZX R8W, byte ptr [R14]
• Chain instructions: MOVSX R8, R8W;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8
• Chain latency: 5
• Code:

     0:	66 45 0f b6 06       	movzx  r8w,BYTE PTR [r14]
     5:	4d 0f bf c0          	movsx  r8,r8w
     9:	4d 31 c6             	xor    r14,r8
     c:	4d 31 c6             	xor    r14,r8
     f:	4d 31 c6             	xor    r14,r8
    12:	4d 31 c6             	xor    r14,r8
  

• Show nanoBench command
• Results:
  • Instructions retired: 6.0
  • Core cycles: 11.0
  • Reference cycles: 9.65
  • UOPS_EXECUTED.CORE: 7.0

Experiment 1

• Instruction: MOVZX R8W, byte ptr [R14+R13]
• Chain instructions: MOVSX R8, R8W;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8
• Chain latency: 5
• Code:

     0:	66 47 0f b6 04 2e    	movzx  r8w,BYTE PTR [r14+r13*1]
     6:	4d 0f bf c0          	movsx  r8,r8w
     a:	4d 31 c5             	xor    r13,r8
     d:	4d 31 c5             	xor    r13,r8
    10:	4d 31 c5             	xor    r13,r8
    13:	4d 31 c5             	xor    r13,r8
  

• Init:

  XOR R13, R13

• Show nanoBench command
• Results:
  • Instructions retired: 6.0
  • Core cycles: 11.0
  • Reference cycles: 9.65
  • UOPS_EXECUTED.CORE: 7.0

Experiment 1

• Instruction: MOVZX R8W, byte ptr [R14]
• Chain instructions: MOVSX R12, R8W;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;mov [R14], R12B
• Chain latency: ≥5
• Code:

     0:	66 45 0f b6 06       	movzx  r8w,BYTE PTR [r14]
     5:	4d 0f bf e0          	movsx  r12,r8w
     9:	4d 63 e4             	movsxd r12,r12d
     c:	4d 63 e4             	movsxd r12,r12d
     f:	4d 63 e4             	movsxd r12,r12d
    12:	45 88 26             	mov    BYTE PTR [r14],r12b
  

• Show nanoBench command
• Results:
  • Instructions retired: 6.0
  • Core cycles: 10.0
  • Reference cycles: 8.96
  • UOPS_EXECUTED.CORE: 8.0