PEXT (R32, R32, M32)

Experiment 1

Instruction: PEXT R9D, R8D, dword ptr [R14]
Chain instruction: PEXT R8D, R9D, dword ptr [R14]

Code:

   0:	c4 42 3a f5 0e       	pext   r9d,r8d,DWORD PTR [r14]
   5:	c4 42 32 f5 06       	pext   r8d,r9d,DWORD PTR [r14]

Init:

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

Show nanoBench command
Results:
- MPERF: 38.58
- APERF: 46.0
- UOPS: 14.0

Experiment 2

Instruction: PEXT R9D, R8D, dword ptr [R14]
Chain instructions: MOVSX R8, R9D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D;MOVSX R8, R8D
Chain latency: 47

Code:

   0:	c4 42 3a f5 0e       	pext   r9d,r8d,DWORD PTR [r14]
   5:	4d 63 c1             	movsxd r8,r9d
   8:	4d 63 c0             	movsxd r8,r8d
   b:	4d 63 c0             	movsxd r8,r8d
   e:	4d 63 c0             	movsxd r8,r8d
  11:	4d 63 c0             	movsxd r8,r8d
  14:	4d 63 c0             	movsxd r8,r8d
  17:	4d 63 c0             	movsxd r8,r8d
  1a:	4d 63 c0             	movsxd r8,r8d
  1d:	4d 63 c0             	movsxd r8,r8d
  20:	4d 63 c0             	movsxd r8,r8d
  23:	4d 63 c0             	movsxd r8,r8d
  26:	4d 63 c0             	movsxd r8,r8d
  29:	4d 63 c0             	movsxd r8,r8d
  2c:	4d 63 c0             	movsxd r8,r8d
  2f:	4d 63 c0             	movsxd r8,r8d
  32:	4d 63 c0             	movsxd r8,r8d
  35:	4d 63 c0             	movsxd r8,r8d
  38:	4d 63 c0             	movsxd r8,r8d
  3b:	4d 63 c0             	movsxd r8,r8d
  3e:	4d 63 c0             	movsxd r8,r8d
  41:	4d 63 c0             	movsxd r8,r8d
  44:	4d 63 c0             	movsxd r8,r8d
  47:	4d 63 c0             	movsxd r8,r8d
  4a:	4d 63 c0             	movsxd r8,r8d
  4d:	4d 63 c0             	movsxd r8,r8d
  50:	4d 63 c0             	movsxd r8,r8d
  53:	4d 63 c0             	movsxd r8,r8d
  56:	4d 63 c0             	movsxd r8,r8d
  59:	4d 63 c0             	movsxd r8,r8d
  5c:	4d 63 c0             	movsxd r8,r8d
  5f:	4d 63 c0             	movsxd r8,r8d
  62:	4d 63 c0             	movsxd r8,r8d
  65:	4d 63 c0             	movsxd r8,r8d
  68:	4d 63 c0             	movsxd r8,r8d
  6b:	4d 63 c0             	movsxd r8,r8d
  6e:	4d 63 c0             	movsxd r8,r8d
  71:	4d 63 c0             	movsxd r8,r8d
  74:	4d 63 c0             	movsxd r8,r8d
  77:	4d 63 c0             	movsxd r8,r8d
  7a:	4d 63 c0             	movsxd r8,r8d
  7d:	4d 63 c0             	movsxd r8,r8d
  80:	4d 63 c0             	movsxd r8,r8d
  83:	4d 63 c0             	movsxd r8,r8d
  86:	4d 63 c0             	movsxd r8,r8d
  89:	4d 63 c0             	movsxd r8,r8d
  8c:	4d 63 c0             	movsxd r8,r8d
  8f:	4d 63 c0             	movsxd r8,r8d

Init:

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

Show nanoBench command
Results:
- MPERF: 29.4
- APERF: 34.8
- UOPS: 54.0

Latency operand 3 → 1 (address, base register): 22

Experiment 1

Instruction: PEXT R8D, R9D, dword ptr [R14]
Chain instructions: MOVSX R8, R8D;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8
Chain latency: 47

Code:

   0:	c4 42 32 f5 06       	pext   r8d,r9d,DWORD PTR [r14]
   5:	4d 63 c0             	movsxd r8,r8d
   8:	4d 31 c6             	xor    r14,r8
   b:	4d 31 c6             	xor    r14,r8
   e:	4d 31 c6             	xor    r14,r8
  11:	4d 31 c6             	xor    r14,r8
  14:	4d 31 c6             	xor    r14,r8
  17:	4d 31 c6             	xor    r14,r8
  1a:	4d 31 c6             	xor    r14,r8
  1d:	4d 31 c6             	xor    r14,r8
  20:	4d 31 c6             	xor    r14,r8
  23:	4d 31 c6             	xor    r14,r8
  26:	4d 31 c6             	xor    r14,r8
  29:	4d 31 c6             	xor    r14,r8
  2c:	4d 31 c6             	xor    r14,r8
  2f:	4d 31 c6             	xor    r14,r8
  32:	4d 31 c6             	xor    r14,r8
  35:	4d 31 c6             	xor    r14,r8
  38:	4d 31 c6             	xor    r14,r8
  3b:	4d 31 c6             	xor    r14,r8
  3e:	4d 31 c6             	xor    r14,r8
  41:	4d 31 c6             	xor    r14,r8
  44:	4d 31 c6             	xor    r14,r8
  47:	4d 31 c6             	xor    r14,r8
  4a:	4d 31 c6             	xor    r14,r8
  4d:	4d 31 c6             	xor    r14,r8
  50:	4d 31 c6             	xor    r14,r8
  53:	4d 31 c6             	xor    r14,r8
  56:	4d 31 c6             	xor    r14,r8
  59:	4d 31 c6             	xor    r14,r8
  5c:	4d 31 c6             	xor    r14,r8
  5f:	4d 31 c6             	xor    r14,r8
  62:	4d 31 c6             	xor    r14,r8
  65:	4d 31 c6             	xor    r14,r8
  68:	4d 31 c6             	xor    r14,r8
  6b:	4d 31 c6             	xor    r14,r8
  6e:	4d 31 c6             	xor    r14,r8
  71:	4d 31 c6             	xor    r14,r8
  74:	4d 31 c6             	xor    r14,r8
  77:	4d 31 c6             	xor    r14,r8
  7a:	4d 31 c6             	xor    r14,r8
  7d:	4d 31 c6             	xor    r14,r8
  80:	4d 31 c6             	xor    r14,r8
  83:	4d 31 c6             	xor    r14,r8
  86:	4d 31 c6             	xor    r14,r8
  89:	4d 31 c6             	xor    r14,r8
  8c:	4d 31 c6             	xor    r14,r8
  8f:	4d 31 c6             	xor    r14,r8

Init:

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

Show nanoBench command
Results:
- MPERF: 58.08
- APERF: 69.0
- UOPS: 54.0

Experiment 2 (with clean upper 32 bits)

Instruction: PEXT R8D, R9D, dword ptr [R14]
Chain instructions: MOVSX R8, R8D;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8;XOR R14, R8
Chain latency: 47

Code:

   0:	c4 42 32 f5 06       	pext   r8d,r9d,DWORD PTR [r14]
   5:	4d 63 c0             	movsxd r8,r8d
   8:	4d 31 c6             	xor    r14,r8
   b:	4d 31 c6             	xor    r14,r8
   e:	4d 31 c6             	xor    r14,r8
  11:	4d 31 c6             	xor    r14,r8
  14:	4d 31 c6             	xor    r14,r8
  17:	4d 31 c6             	xor    r14,r8
  1a:	4d 31 c6             	xor    r14,r8
  1d:	4d 31 c6             	xor    r14,r8
  20:	4d 31 c6             	xor    r14,r8
  23:	4d 31 c6             	xor    r14,r8
  26:	4d 31 c6             	xor    r14,r8
  29:	4d 31 c6             	xor    r14,r8
  2c:	4d 31 c6             	xor    r14,r8
  2f:	4d 31 c6             	xor    r14,r8
  32:	4d 31 c6             	xor    r14,r8
  35:	4d 31 c6             	xor    r14,r8
  38:	4d 31 c6             	xor    r14,r8
  3b:	4d 31 c6             	xor    r14,r8
  3e:	4d 31 c6             	xor    r14,r8
  41:	4d 31 c6             	xor    r14,r8
  44:	4d 31 c6             	xor    r14,r8
  47:	4d 31 c6             	xor    r14,r8
  4a:	4d 31 c6             	xor    r14,r8
  4d:	4d 31 c6             	xor    r14,r8
  50:	4d 31 c6             	xor    r14,r8
  53:	4d 31 c6             	xor    r14,r8
  56:	4d 31 c6             	xor    r14,r8
  59:	4d 31 c6             	xor    r14,r8
  5c:	4d 31 c6             	xor    r14,r8
  5f:	4d 31 c6             	xor    r14,r8
  62:	4d 31 c6             	xor    r14,r8
  65:	4d 31 c6             	xor    r14,r8
  68:	4d 31 c6             	xor    r14,r8
  6b:	4d 31 c6             	xor    r14,r8
  6e:	4d 31 c6             	xor    r14,r8
  71:	4d 31 c6             	xor    r14,r8
  74:	4d 31 c6             	xor    r14,r8
  77:	4d 31 c6             	xor    r14,r8
  7a:	4d 31 c6             	xor    r14,r8
  7d:	4d 31 c6             	xor    r14,r8
  80:	4d 31 c6             	xor    r14,r8
  83:	4d 31 c6             	xor    r14,r8
  86:	4d 31 c6             	xor    r14,r8
  89:	4d 31 c6             	xor    r14,r8
  8c:	4d 31 c6             	xor    r14,r8
  8f:	4d 31 c6             	xor    r14,r8

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV R9D, R9D;

Show nanoBench command
Results:
- MPERF: 57.78
- APERF: 69.0
- UOPS: 54.0

Latency operand 3 → 1 (address, index register): 22

Experiment 1

Instruction: PEXT R8D, R9D, dword ptr [R14+R13]
Chain instructions: MOVSX R8, R8D;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8
Chain latency: 47

Code:

   0:	c4 02 32 f5 04 2e    	pext   r8d,r9d,DWORD PTR [r14+r13*1]
   6:	4d 63 c0             	movsxd r8,r8d
   9:	4d 31 c5             	xor    r13,r8
   c:	4d 31 c5             	xor    r13,r8
   f:	4d 31 c5             	xor    r13,r8
  12:	4d 31 c5             	xor    r13,r8
  15:	4d 31 c5             	xor    r13,r8
  18:	4d 31 c5             	xor    r13,r8
  1b:	4d 31 c5             	xor    r13,r8
  1e:	4d 31 c5             	xor    r13,r8
  21:	4d 31 c5             	xor    r13,r8
  24:	4d 31 c5             	xor    r13,r8
  27:	4d 31 c5             	xor    r13,r8
  2a:	4d 31 c5             	xor    r13,r8
  2d:	4d 31 c5             	xor    r13,r8
  30:	4d 31 c5             	xor    r13,r8
  33:	4d 31 c5             	xor    r13,r8
  36:	4d 31 c5             	xor    r13,r8
  39:	4d 31 c5             	xor    r13,r8
  3c:	4d 31 c5             	xor    r13,r8
  3f:	4d 31 c5             	xor    r13,r8
  42:	4d 31 c5             	xor    r13,r8
  45:	4d 31 c5             	xor    r13,r8
  48:	4d 31 c5             	xor    r13,r8
  4b:	4d 31 c5             	xor    r13,r8
  4e:	4d 31 c5             	xor    r13,r8
  51:	4d 31 c5             	xor    r13,r8
  54:	4d 31 c5             	xor    r13,r8
  57:	4d 31 c5             	xor    r13,r8
  5a:	4d 31 c5             	xor    r13,r8
  5d:	4d 31 c5             	xor    r13,r8
  60:	4d 31 c5             	xor    r13,r8
  63:	4d 31 c5             	xor    r13,r8
  66:	4d 31 c5             	xor    r13,r8
  69:	4d 31 c5             	xor    r13,r8
  6c:	4d 31 c5             	xor    r13,r8
  6f:	4d 31 c5             	xor    r13,r8
  72:	4d 31 c5             	xor    r13,r8
  75:	4d 31 c5             	xor    r13,r8
  78:	4d 31 c5             	xor    r13,r8
  7b:	4d 31 c5             	xor    r13,r8
  7e:	4d 31 c5             	xor    r13,r8
  81:	4d 31 c5             	xor    r13,r8
  84:	4d 31 c5             	xor    r13,r8
  87:	4d 31 c5             	xor    r13,r8
  8a:	4d 31 c5             	xor    r13,r8
  8d:	4d 31 c5             	xor    r13,r8
  90:	4d 31 c5             	xor    r13,r8

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
XOR R13, R13

Show nanoBench command
Results:
- MPERF: 58.08
- APERF: 69.0
- UOPS: 54.0

Experiment 2 (with clean upper 32 bits)

Instruction: PEXT R8D, R9D, dword ptr [R14+R13]
Chain instructions: MOVSX R8, R8D;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8;XOR R13, R8
Chain latency: 47

Code:

   0:	c4 02 32 f5 04 2e    	pext   r8d,r9d,DWORD PTR [r14+r13*1]
   6:	4d 63 c0             	movsxd r8,r8d
   9:	4d 31 c5             	xor    r13,r8
   c:	4d 31 c5             	xor    r13,r8
   f:	4d 31 c5             	xor    r13,r8
  12:	4d 31 c5             	xor    r13,r8
  15:	4d 31 c5             	xor    r13,r8
  18:	4d 31 c5             	xor    r13,r8
  1b:	4d 31 c5             	xor    r13,r8
  1e:	4d 31 c5             	xor    r13,r8
  21:	4d 31 c5             	xor    r13,r8
  24:	4d 31 c5             	xor    r13,r8
  27:	4d 31 c5             	xor    r13,r8
  2a:	4d 31 c5             	xor    r13,r8
  2d:	4d 31 c5             	xor    r13,r8
  30:	4d 31 c5             	xor    r13,r8
  33:	4d 31 c5             	xor    r13,r8
  36:	4d 31 c5             	xor    r13,r8
  39:	4d 31 c5             	xor    r13,r8
  3c:	4d 31 c5             	xor    r13,r8
  3f:	4d 31 c5             	xor    r13,r8
  42:	4d 31 c5             	xor    r13,r8
  45:	4d 31 c5             	xor    r13,r8
  48:	4d 31 c5             	xor    r13,r8
  4b:	4d 31 c5             	xor    r13,r8
  4e:	4d 31 c5             	xor    r13,r8
  51:	4d 31 c5             	xor    r13,r8
  54:	4d 31 c5             	xor    r13,r8
  57:	4d 31 c5             	xor    r13,r8
  5a:	4d 31 c5             	xor    r13,r8
  5d:	4d 31 c5             	xor    r13,r8
  60:	4d 31 c5             	xor    r13,r8
  63:	4d 31 c5             	xor    r13,r8
  66:	4d 31 c5             	xor    r13,r8
  69:	4d 31 c5             	xor    r13,r8
  6c:	4d 31 c5             	xor    r13,r8
  6f:	4d 31 c5             	xor    r13,r8
  72:	4d 31 c5             	xor    r13,r8
  75:	4d 31 c5             	xor    r13,r8
  78:	4d 31 c5             	xor    r13,r8
  7b:	4d 31 c5             	xor    r13,r8
  7e:	4d 31 c5             	xor    r13,r8
  81:	4d 31 c5             	xor    r13,r8
  84:	4d 31 c5             	xor    r13,r8
  87:	4d 31 c5             	xor    r13,r8
  8a:	4d 31 c5             	xor    r13,r8
  8d:	4d 31 c5             	xor    r13,r8
  90:	4d 31 c5             	xor    r13,r8

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
XOR R13, R13;
MOV R9D, R9D;

Show nanoBench command
Results:
- MPERF: 58.08
- APERF: 69.0
- UOPS: 54.0

Latency operand 3 → 1 (memory): ≤25

Experiment 1

Instruction: PEXT R8D, R9D, dword ptr [R14]
Chain instructions: MOVSX R12, R8D;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;mov [R14], R12D
Chain latency: ≥46

Code:

   0:	c4 42 32 f5 06       	pext   r8d,r9d,DWORD PTR [r14]
   5:	4d 63 e0             	movsxd r12,r8d
   8:	4d 63 e4             	movsxd r12,r12d
   b:	4d 63 e4             	movsxd r12,r12d
   e:	4d 63 e4             	movsxd r12,r12d
  11:	4d 63 e4             	movsxd r12,r12d
  14:	4d 63 e4             	movsxd r12,r12d
  17:	4d 63 e4             	movsxd r12,r12d
  1a:	4d 63 e4             	movsxd r12,r12d
  1d:	4d 63 e4             	movsxd r12,r12d
  20:	4d 63 e4             	movsxd r12,r12d
  23:	4d 63 e4             	movsxd r12,r12d
  26:	4d 63 e4             	movsxd r12,r12d
  29:	4d 63 e4             	movsxd r12,r12d
  2c:	4d 63 e4             	movsxd r12,r12d
  2f:	4d 63 e4             	movsxd r12,r12d
  32:	4d 63 e4             	movsxd r12,r12d
  35:	4d 63 e4             	movsxd r12,r12d
  38:	4d 63 e4             	movsxd r12,r12d
  3b:	4d 63 e4             	movsxd r12,r12d
  3e:	4d 63 e4             	movsxd r12,r12d
  41:	4d 63 e4             	movsxd r12,r12d
  44:	4d 63 e4             	movsxd r12,r12d
  47:	4d 63 e4             	movsxd r12,r12d
  4a:	4d 63 e4             	movsxd r12,r12d
  4d:	4d 63 e4             	movsxd r12,r12d
  50:	4d 63 e4             	movsxd r12,r12d
  53:	4d 63 e4             	movsxd r12,r12d
  56:	4d 63 e4             	movsxd r12,r12d
  59:	4d 63 e4             	movsxd r12,r12d
  5c:	4d 63 e4             	movsxd r12,r12d
  5f:	4d 63 e4             	movsxd r12,r12d
  62:	4d 63 e4             	movsxd r12,r12d
  65:	4d 63 e4             	movsxd r12,r12d
  68:	4d 63 e4             	movsxd r12,r12d
  6b:	4d 63 e4             	movsxd r12,r12d
  6e:	4d 63 e4             	movsxd r12,r12d
  71:	4d 63 e4             	movsxd r12,r12d
  74:	4d 63 e4             	movsxd r12,r12d
  77:	4d 63 e4             	movsxd r12,r12d
  7a:	4d 63 e4             	movsxd r12,r12d
  7d:	4d 63 e4             	movsxd r12,r12d
  80:	4d 63 e4             	movsxd r12,r12d
  83:	4d 63 e4             	movsxd r12,r12d
  86:	4d 63 e4             	movsxd r12,r12d
  89:	4d 63 e4             	movsxd r12,r12d
  8c:	4d 63 e4             	movsxd r12,r12d
  8f:	45 89 26             	mov    DWORD PTR [r14],r12d

Init:

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

Show nanoBench command
Results:
- MPERF: 59.76
- APERF: 71.0
- UOPS: 54.0

Experiment 2 (with clean upper 32 bits)

Instruction: PEXT R8D, R9D, dword ptr [R14]
Chain instructions: MOVSX R12, R8D;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;MOVSX R12, R12d;mov [R14], R12D
Chain latency: ≥46

Code:

   0:	c4 42 32 f5 06       	pext   r8d,r9d,DWORD PTR [r14]
   5:	4d 63 e0             	movsxd r12,r8d
   8:	4d 63 e4             	movsxd r12,r12d
   b:	4d 63 e4             	movsxd r12,r12d
   e:	4d 63 e4             	movsxd r12,r12d
  11:	4d 63 e4             	movsxd r12,r12d
  14:	4d 63 e4             	movsxd r12,r12d
  17:	4d 63 e4             	movsxd r12,r12d
  1a:	4d 63 e4             	movsxd r12,r12d
  1d:	4d 63 e4             	movsxd r12,r12d
  20:	4d 63 e4             	movsxd r12,r12d
  23:	4d 63 e4             	movsxd r12,r12d
  26:	4d 63 e4             	movsxd r12,r12d
  29:	4d 63 e4             	movsxd r12,r12d
  2c:	4d 63 e4             	movsxd r12,r12d
  2f:	4d 63 e4             	movsxd r12,r12d
  32:	4d 63 e4             	movsxd r12,r12d
  35:	4d 63 e4             	movsxd r12,r12d
  38:	4d 63 e4             	movsxd r12,r12d
  3b:	4d 63 e4             	movsxd r12,r12d
  3e:	4d 63 e4             	movsxd r12,r12d
  41:	4d 63 e4             	movsxd r12,r12d
  44:	4d 63 e4             	movsxd r12,r12d
  47:	4d 63 e4             	movsxd r12,r12d
  4a:	4d 63 e4             	movsxd r12,r12d
  4d:	4d 63 e4             	movsxd r12,r12d
  50:	4d 63 e4             	movsxd r12,r12d
  53:	4d 63 e4             	movsxd r12,r12d
  56:	4d 63 e4             	movsxd r12,r12d
  59:	4d 63 e4             	movsxd r12,r12d
  5c:	4d 63 e4             	movsxd r12,r12d
  5f:	4d 63 e4             	movsxd r12,r12d
  62:	4d 63 e4             	movsxd r12,r12d
  65:	4d 63 e4             	movsxd r12,r12d
  68:	4d 63 e4             	movsxd r12,r12d
  6b:	4d 63 e4             	movsxd r12,r12d
  6e:	4d 63 e4             	movsxd r12,r12d
  71:	4d 63 e4             	movsxd r12,r12d
  74:	4d 63 e4             	movsxd r12,r12d
  77:	4d 63 e4             	movsxd r12,r12d
  7a:	4d 63 e4             	movsxd r12,r12d
  7d:	4d 63 e4             	movsxd r12,r12d
  80:	4d 63 e4             	movsxd r12,r12d
  83:	4d 63 e4             	movsxd r12,r12d
  86:	4d 63 e4             	movsxd r12,r12d
  89:	4d 63 e4             	movsxd r12,r12d
  8c:	4d 63 e4             	movsxd r12,r12d
  8f:	45 89 26             	mov    DWORD PTR [r14],r12d

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV R9D, R9D;

Show nanoBench command
Results:
- MPERF: 59.88
- APERF: 71.0
- UOPS: 54.0

PEXT (R32, R32, M32) - Latency

Operands

Latency operand 2 → 1: 0

Latency operand 3 → 1 (address, base register): 22

Latency operand 3 → 1 (address, index register): 22

Latency operand 3 → 1 (memory): ≤25

Latency operand 2 → 1: 0

Experiment 1

Experiment 2

Latency operand 3 → 1 (address, base register): 22

Experiment 1

Experiment 2 (with clean upper 32 bits)

Latency operand 3 → 1 (address, index register): 22

Experiment 1

Experiment 2 (with clean upper 32 bits)

Latency operand 3 → 1 (memory): ≤25

Experiment 1

Experiment 2 (with clean upper 32 bits)