VDIVPS (YMM, YMM, YMM)

Experiment 1

Instruction: VDIVPS YMM0, YMM2, YMM4
Chain instructions: VORPS YMM2, YMM0, YMM0; VORPS YMM2, YMM2, YMM1; VANDPS YMM2, YMM2, YMM1; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2;
Chain latency: ≥21

Code:

   0:	c5 ec 5e c4          	vdivps ymm0,ymm2,ymm4
   4:	c5 fc 56 d0          	vorps  ymm2,ymm0,ymm0
   8:	c5 ec 56 d1          	vorps  ymm2,ymm2,ymm1
   c:	c5 ec 54 d1          	vandps ymm2,ymm2,ymm1
  10:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  14:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  18:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  1c:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  20:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  24:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  28:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  2c:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  30:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  34:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  38:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  3c:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  40:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  44:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  48:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  4c:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  50:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  54:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x54ed392654ed3926;
MOV RBX, 0x3f99f4c03f99f4c0;
MOV [R14+0], RBX;
MOV [R14+8], RBX;
MOV [R14+16], RBX;
MOV [R14+24], RBX;
MOV [R14+64], RAX;
MOV [R14+72], RAX;
MOV [R14+80], RAX;
MOV [R14+88], RAX;
VMOVUPS YMM1, [R14+64];
VMOVUPS YMM2, [R14+64];
VMOVUPS YMM3, [R14];
VMOVUPS YMM4, [R14]

Show nanoBench command
Results:
- Instructions retired: 22.0
- Core cycles: 42.35
- Reference cycles: 42.29
- UOPS_EXECUTED.CORE: 24.0

Experiment 2

Instruction: VDIVPS YMM2, YMM2, YMM4

Code:

   0:	c5 ec 5e d4          	vdivps ymm2,ymm2,ymm4

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x3f8000003f800000;
MOV RBX, 0x3f8000003f800000;
MOV [R14+0], RBX;
MOV [R14+8], RBX;
MOV [R14+16], RBX;
MOV [R14+24], RBX;
MOV [R14+64], RAX;
MOV [R14+72], RAX;
MOV [R14+80], RAX;
MOV [R14+88], RAX;
VMOVUPS YMM1, [R14+64];
VMOVUPS YMM2, [R14+64];
VMOVUPS YMM3, [R14];
VMOVUPS YMM4, [R14]

Show nanoBench command
Results:
- Instructions retired: 1.0
- Core cycles: 18.15
- Reference cycles: 18.13
- UOPS_EXECUTED.CORE: 3.0

Experiment 3

Instruction: VDIVPS YMM0, YMM2, YMM4
Chain instructions: VORPS YMM2, YMM0, YMM0; VORPS YMM2, YMM2, YMM1; VANDPS YMM2, YMM2, YMM1; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2; VORPS YMM2, YMM2, YMM2;
Chain latency: ≥21

Code:

   0:	c5 ec 5e c4          	vdivps ymm0,ymm2,ymm4
   4:	c5 fc 56 d0          	vorps  ymm2,ymm0,ymm0
   8:	c5 ec 56 d1          	vorps  ymm2,ymm2,ymm1
   c:	c5 ec 54 d1          	vandps ymm2,ymm2,ymm1
  10:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  14:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  18:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  1c:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  20:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  24:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  28:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  2c:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  30:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  34:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  38:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  3c:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  40:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  44:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  48:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  4c:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  50:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2
  54:	c5 ec 56 d2          	vorps  ymm2,ymm2,ymm2

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x3f8000003f800000;
MOV RBX, 0x3f8000003f800000;
MOV [R14+0], RBX;
MOV [R14+8], RBX;
MOV [R14+16], RBX;
MOV [R14+24], RBX;
MOV [R14+64], RAX;
MOV [R14+72], RAX;
MOV [R14+80], RAX;
MOV [R14+88], RAX;
VMOVUPS YMM1, [R14+64];
VMOVUPS YMM2, [R14+64];
VMOVUPS YMM3, [R14];
VMOVUPS YMM4, [R14]

Show nanoBench command
Results:
- Instructions retired: 22.0
- Core cycles: 39.36
- Reference cycles: 39.47
- UOPS_EXECUTED.CORE: 24.0

Experiment 1

Instruction: VDIVPS YMM0, YMM2, YMM4
Chain instructions: VORPS YMM4, YMM0, YMM0; VORPS YMM4, YMM4, YMM3; VANDPS YMM4, YMM4, YMM3; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4;
Chain latency: ≥21

Code:

   0:	c5 ec 5e c4          	vdivps ymm0,ymm2,ymm4
   4:	c5 fc 56 e0          	vorps  ymm4,ymm0,ymm0
   8:	c5 dc 56 e3          	vorps  ymm4,ymm4,ymm3
   c:	c5 dc 54 e3          	vandps ymm4,ymm4,ymm3
  10:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  14:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  18:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  1c:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  20:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  24:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  28:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  2c:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  30:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  34:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  38:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  3c:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  40:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  44:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  48:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  4c:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  50:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  54:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x54ed392654ed3926;
MOV RBX, 0x3f99f4c03f99f4c0;
MOV [R14+0], RBX;
MOV [R14+8], RBX;
MOV [R14+16], RBX;
MOV [R14+24], RBX;
MOV [R14+64], RAX;
MOV [R14+72], RAX;
MOV [R14+80], RAX;
MOV [R14+88], RAX;
VMOVUPS YMM1, [R14+64];
VMOVUPS YMM2, [R14+64];
VMOVUPS YMM3, [R14];
VMOVUPS YMM4, [R14]

Show nanoBench command
Results:
- Instructions retired: 22.0
- Core cycles: 42.4
- Reference cycles: 42.45
- UOPS_EXECUTED.CORE: 24.0

Experiment 2

Instruction: VDIVPS YMM4, YMM2, YMM4

Code:

   0:	c5 ec 5e e4          	vdivps ymm4,ymm2,ymm4

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x3f8000003f800000;
MOV RBX, 0x3f8000003f800000;
MOV [R14+0], RBX;
MOV [R14+8], RBX;
MOV [R14+16], RBX;
MOV [R14+24], RBX;
MOV [R14+64], RAX;
MOV [R14+72], RAX;
MOV [R14+80], RAX;
MOV [R14+88], RAX;
VMOVUPS YMM1, [R14+64];
VMOVUPS YMM2, [R14+64];
VMOVUPS YMM3, [R14];
VMOVUPS YMM4, [R14]

Show nanoBench command
Results:
- Instructions retired: 1.0
- Core cycles: 18.14
- Reference cycles: 18.24
- UOPS_EXECUTED.CORE: 3.0

Experiment 3

Instruction: VDIVPS YMM0, YMM2, YMM4
Chain instructions: VORPS YMM4, YMM0, YMM0; VORPS YMM4, YMM4, YMM3; VANDPS YMM4, YMM4, YMM3; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4; VORPS YMM4, YMM4, YMM4;
Chain latency: ≥21

Code:

   0:	c5 ec 5e c4          	vdivps ymm0,ymm2,ymm4
   4:	c5 fc 56 e0          	vorps  ymm4,ymm0,ymm0
   8:	c5 dc 56 e3          	vorps  ymm4,ymm4,ymm3
   c:	c5 dc 54 e3          	vandps ymm4,ymm4,ymm3
  10:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  14:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  18:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  1c:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  20:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  24:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  28:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  2c:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  30:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  34:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  38:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  3c:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  40:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  44:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  48:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  4c:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  50:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4
  54:	c5 dc 56 e4          	vorps  ymm4,ymm4,ymm4

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x3f8000003f800000;
MOV RBX, 0x3f8000003f800000;
MOV [R14+0], RBX;
MOV [R14+8], RBX;
MOV [R14+16], RBX;
MOV [R14+24], RBX;
MOV [R14+64], RAX;
MOV [R14+72], RAX;
MOV [R14+80], RAX;
MOV [R14+88], RAX;
VMOVUPS YMM1, [R14+64];
VMOVUPS YMM2, [R14+64];
VMOVUPS YMM3, [R14];
VMOVUPS YMM4, [R14]

Show nanoBench command
Results:
- Instructions retired: 22.0
- Core cycles: 39.35
- Reference cycles: 39.25
- UOPS_EXECUTED.CORE: 24.0

VDIVPS (YMM, YMM, YMM) - Latency

Operands

Latency operand 2 → 1: ≤18

Latency operand 3 → 1: ≤18

Latency operand 2 → 1: ≤18

Experiment 1

Experiment 2

Experiment 3

Latency operand 3 → 1: ≤18

Experiment 1

Experiment 2

Experiment 3