DIV (R8l)

Experiment 1 (fast division)

Instruction: DIV BL
Chain instructions: AND BL, AL; AND BL, CL; OR BL, CL; OR BL, BL; OR BL, BL; OR BL, BL; OR BL, BL; OR BL, BL; OR BL, BL; OR BL, BL; OR BL, BL; OR BL, BL; OR BL, BL; ; MOV AX, R8W
Chain latency: 13

Code:

   0:	f6 f3                	div    bl
   2:	20 c3                	and    bl,al
   4:	20 cb                	and    bl,cl
   6:	08 cb                	or     bl,cl
   8:	08 db                	or     bl,bl
   a:	08 db                	or     bl,bl
   c:	08 db                	or     bl,bl
   e:	08 db                	or     bl,bl
  10:	08 db                	or     bl,bl
  12:	08 db                	or     bl,bl
  14:	08 db                	or     bl,bl
  16:	08 db                	or     bl,bl
  18:	08 db                	or     bl,bl
  1a:	08 db                	or     bl,bl
  1c:	66 44 89 c0          	mov    ax,r8w

Init:

MOV R8, 0;
MOV R9, 0;
MOV RCX, 1;
MOV BL, CL;
MOV RAX, R8;
MOV RDX, R9

Show nanoBench command
Results:
- Instructions retired: 15.0
- Core cycles: 29.0
- Reference cycles: 24.13
- UOPS_RETIRED.ANY: 18.0

Experiment 2 (slow division)

Instruction: DIV BL
Chain instructions: AND BL, AL; AND BL, CL; OR BL, CL; OR BL, BL; OR BL, BL; OR BL, BL; OR BL, BL; OR BL, BL; OR BL, BL; OR BL, BL; OR BL, BL; OR BL, BL; OR BL, BL; ; MOV AX, R8W
Chain latency: 13

Code:

   0:	f6 f3                	div    bl
   2:	20 c3                	and    bl,al
   4:	20 cb                	and    bl,cl
   6:	08 cb                	or     bl,cl
   8:	08 db                	or     bl,bl
   a:	08 db                	or     bl,bl
   c:	08 db                	or     bl,bl
   e:	08 db                	or     bl,bl
  10:	08 db                	or     bl,bl
  12:	08 db                	or     bl,bl
  14:	08 db                	or     bl,bl
  16:	08 db                	or     bl,bl
  18:	08 db                	or     bl,bl
  1a:	08 db                	or     bl,bl
  1c:	66 44 89 c0          	mov    ax,r8w

Init:

MOV R8, 13057;
MOV R9, 0;
MOV RCX, 123;
MOV BL, CL;
MOV RAX, R8;
MOV RDX, R9

Show nanoBench command
Results:
- Instructions retired: 15.0
- Core cycles: 33.0
- Reference cycles: 27.53
- UOPS_RETIRED.ANY: 18.0

Experiment 1 (fast division)

Instruction: DIV BL
Chain instructions: AND AX, R8W; OR AX, R8W; OR AX, AX; OR AX, AX; OR AX, AX; OR AX, AX; OR AX, AX; OR AX, AX; OR AX, AX; OR AX, AX; OR AX, AX; OR AX, AX; ;
Chain latency: 12

Code:

   0:	f6 f3                	div    bl
   2:	66 44 21 c0          	and    ax,r8w
   6:	66 44 09 c0          	or     ax,r8w
   a:	66 09 c0             	or     ax,ax
   d:	66 09 c0             	or     ax,ax
  10:	66 09 c0             	or     ax,ax
  13:	66 09 c0             	or     ax,ax
  16:	66 09 c0             	or     ax,ax
  19:	66 09 c0             	or     ax,ax
  1c:	66 09 c0             	or     ax,ax
  1f:	66 09 c0             	or     ax,ax
  22:	66 09 c0             	or     ax,ax
  25:	66 09 c0             	or     ax,ax

Init:

MOV R8, 0;
MOV R9, 0;
MOV RCX, 1;
MOV BL, CL;
MOV RAX, R8;
MOV RDX, R9

Show nanoBench command
Results:
- Instructions retired: 13.0
- Core cycles: 27.0
- Reference cycles: 22.43
- UOPS_RETIRED.ANY: 16.0

Experiment 2 (slow division)

Instruction: DIV BL
Chain instructions: AND AX, R8W; OR AX, R8W; OR AX, AX; OR AX, AX; OR AX, AX; OR AX, AX; OR AX, AX; OR AX, AX; OR AX, AX; OR AX, AX; OR AX, AX; OR AX, AX; ;
Chain latency: 12

Code:

   0:	f6 f3                	div    bl
   2:	66 44 21 c0          	and    ax,r8w
   6:	66 44 09 c0          	or     ax,r8w
   a:	66 09 c0             	or     ax,ax
   d:	66 09 c0             	or     ax,ax
  10:	66 09 c0             	or     ax,ax
  13:	66 09 c0             	or     ax,ax
  16:	66 09 c0             	or     ax,ax
  19:	66 09 c0             	or     ax,ax
  1c:	66 09 c0             	or     ax,ax
  1f:	66 09 c0             	or     ax,ax
  22:	66 09 c0             	or     ax,ax
  25:	66 09 c0             	or     ax,ax

Init:

MOV R8, 13057;
MOV R9, 0;
MOV RCX, 123;
MOV BL, CL;
MOV RAX, R8;
MOV RDX, R9

Show nanoBench command
Results:
- Instructions retired: 13.0
- Core cycles: 31.0
- Reference cycles: 26.0
- UOPS_RETIRED.ANY: 16.0

DIV (R8l) - Latency

Operands

Latency operand 1 → 2: 16 ≤ lat ≤ 20

Latency operand 2 → 2: 15 ≤ lat ≤ 19

Latency operand 1 → 2: 16 ≤ lat ≤ 20

Experiment 1 (fast division)

Experiment 2 (slow division)

Latency operand 2 → 2: 15 ≤ lat ≤ 19

Experiment 1 (fast division)

Experiment 2 (slow division)