KMOVW (K, M16) - Throughput and Uops

With unroll_count=500 and no inner loop

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 1.0
  • Core cycles: 1.0
  • Reference cycles: 0.67
  • UOPS_EXECUTED.THREAD: 2.0
  • RETIRE_SLOTS: 3.0
  • UOPS_MITE: 3.0
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 0.5
  • UOPS_PORT_3: 0.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 1.0
  • UOPS_PORT_6: 0.0
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 1.0

With unroll_count=500, no inner loop, and 1 NOP

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
     5:	90                   	nop
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 2.0
  • Core cycles: 1.0
  • Reference cycles: 0.8
  • UOPS_EXECUTED.THREAD: 2.0
  • RETIRE_SLOTS: 4.0
  • UOPS_MITE: 4.0
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 0.5
  • UOPS_PORT_3: 0.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 1.0
  • UOPS_PORT_6: 0.0
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 1.0

With unroll_count=500, no inner loop, and 2 NOPs

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
     5:	90                   	nop
     6:	90                   	nop
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 3.0
  • Core cycles: 1.25
  • Reference cycles: 0.83
  • UOPS_EXECUTED.THREAD: 2.0
  • RETIRE_SLOTS: 5.0
  • UOPS_MITE: 5.0
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 0.5
  • UOPS_PORT_3: 0.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 1.0
  • UOPS_PORT_6: 0.0
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 1.0

With unroll_count=500, no inner loop, and 3 NOPs

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
     5:	90                   	nop
     6:	90                   	nop
     7:	90                   	nop
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 4.0
  • Core cycles: 2.0
  • Reference cycles: 1.47
  • UOPS_EXECUTED.THREAD: 2.0
  • RETIRE_SLOTS: 6.0
  • UOPS_MITE: 6.0
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 0.5
  • UOPS_PORT_3: 0.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 1.0
  • UOPS_PORT_6: 0.0
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 2.0

With loop_count=1000 and unroll_count=10

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 1.2
  • Core cycles: 1.0
  • Reference cycles: 0.72
  • UOPS_EXECUTED.THREAD: 2.1
  • RETIRE_SLOTS: 3.1
  • UOPS_MITE: 0.9
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 0.5
  • UOPS_PORT_3: 0.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 1.0
  • UOPS_PORT_6: 0.1
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.29

With loop_count=1000, unroll_count=10, and padding (long NOPs)

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
     5:	66 66 66 66 66 66 2e 	data16 data16 data16 data16 data16 cs nop WORD PTR [rax+rax*1+0x0]
     c:	0f 1f 84 00 00 00 00 
    13:	00 
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 2.2
  • Core cycles: 1.1
  • Reference cycles: 0.76
  • UOPS_EXECUTED.THREAD: 2.1
  • RETIRE_SLOTS: 4.1
  • UOPS_MITE: 0.0
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 0.5
  • UOPS_PORT_3: 0.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 1.0
  • UOPS_PORT_6: 0.1
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.0

With loop_count=100 and unroll_count=100

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 1.02
  • Core cycles: 1.0
  • Reference cycles: 0.69
  • UOPS_EXECUTED.THREAD: 2.01
  • RETIRE_SLOTS: 3.01
  • UOPS_MITE: 3.01
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 0.5
  • UOPS_PORT_3: 0.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 1.0
  • UOPS_PORT_6: 0.01
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 1.0

With loop_count=100, unroll_count=100, and padding (long NOPs)

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
     5:	66 66 66 66 66 66 2e 	data16 data16 data16 data16 data16 cs nop WORD PTR [rax+rax*1+0x0]
     c:	0f 1f 84 00 00 00 00 
    13:	00 
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 2.02
  • Core cycles: 1.01
  • Reference cycles: 0.69
  • UOPS_EXECUTED.THREAD: 2.01
  • RETIRE_SLOTS: 4.01
  • UOPS_MITE: 0.04
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 0.5
  • UOPS_PORT_3: 0.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 1.0
  • UOPS_PORT_6: 0.01
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.01

With unroll_count=200 and no inner loop

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
     5:	c4 c1 78 90 56 02    	kmovw  k2,WORD PTR [r14+0x2]
     b:	c4 c1 78 90 5e 04    	kmovw  k3,WORD PTR [r14+0x4]
    11:	c4 c1 78 90 66 06    	kmovw  k4,WORD PTR [r14+0x6]
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 4.0
  • Core cycles: 4.0
  • Reference cycles: 2.61
  • UOPS_EXECUTED.THREAD: 8.0
  • RETIRE_SLOTS: 12.0
  • UOPS_MITE: 12.0
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 2.0
  • UOPS_PORT_3: 2.0
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 4.0
  • UOPS_PORT_6: 0.0
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 4.0

With loop_count=1000 and unroll_count=2

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
     5:	c4 c1 78 90 56 02    	kmovw  k2,WORD PTR [r14+0x2]
     b:	c4 c1 78 90 5e 04    	kmovw  k3,WORD PTR [r14+0x4]
    11:	c4 c1 78 90 66 06    	kmovw  k4,WORD PTR [r14+0x6]
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 5.0
  • Core cycles: 4.01
  • Reference cycles: 2.76
  • UOPS_EXECUTED.THREAD: 8.5
  • RETIRE_SLOTS: 12.5
  • UOPS_MITE: 0.23
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 2.0
  • UOPS_PORT_3: 2.0
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 4.0
  • UOPS_PORT_6: 0.5
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.07

With loop_count=1000, unroll_count=2, and padding (long NOPs)

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
     5:	c4 c1 78 90 56 02    	kmovw  k2,WORD PTR [r14+0x2]
     b:	c4 c1 78 90 5e 04    	kmovw  k3,WORD PTR [r14+0x4]
    11:	c4 c1 78 90 66 06    	kmovw  k4,WORD PTR [r14+0x6]
    17:	66 66 66 66 66 66 2e 	data16 data16 data16 data16 data16 cs nop WORD PTR [rax+rax*1+0x0]
    1e:	0f 1f 84 00 00 00 00 
    25:	00 
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 6.0
  • Core cycles: 4.01
  • Reference cycles: 2.78
  • UOPS_EXECUTED.THREAD: 8.5
  • RETIRE_SLOTS: 13.5
  • UOPS_MITE: 0.01
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 2.0
  • UOPS_PORT_3: 2.0
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 4.0
  • UOPS_PORT_6: 0.5
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.01

With loop_count=100 and unroll_count=20

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
     5:	c4 c1 78 90 56 02    	kmovw  k2,WORD PTR [r14+0x2]
     b:	c4 c1 78 90 5e 04    	kmovw  k3,WORD PTR [r14+0x4]
    11:	c4 c1 78 90 66 06    	kmovw  k4,WORD PTR [r14+0x6]
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 4.1
  • Core cycles: 4.01
  • Reference cycles: 2.78
  • UOPS_EXECUTED.THREAD: 8.05
  • RETIRE_SLOTS: 12.05
  • UOPS_MITE: 0.12
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 2.0
  • UOPS_PORT_3: 2.0
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 4.0
  • UOPS_PORT_6: 0.05
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.04

With loop_count=100, unroll_count=20, and padding (long NOPs)

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
     5:	c4 c1 78 90 56 02    	kmovw  k2,WORD PTR [r14+0x2]
     b:	c4 c1 78 90 5e 04    	kmovw  k3,WORD PTR [r14+0x4]
    11:	c4 c1 78 90 66 06    	kmovw  k4,WORD PTR [r14+0x6]
    17:	66 66 66 66 66 66 2e 	data16 data16 data16 data16 data16 cs nop WORD PTR [rax+rax*1+0x0]
    1e:	0f 1f 84 00 00 00 00 
    25:	00 
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 5.1
  • Core cycles: 4.01
  • Reference cycles: 2.78
  • UOPS_EXECUTED.THREAD: 8.05
  • RETIRE_SLOTS: 13.05
  • UOPS_MITE: 0.13
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 2.0
  • UOPS_PORT_3: 2.0
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 4.0
  • UOPS_PORT_6: 0.05
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.04

With unroll_count=100 and no inner loop

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
     5:	c4 c1 78 90 56 02    	kmovw  k2,WORD PTR [r14+0x2]
     b:	c4 c1 78 90 5e 04    	kmovw  k3,WORD PTR [r14+0x4]
    11:	c4 c1 78 90 66 06    	kmovw  k4,WORD PTR [r14+0x6]
    17:	c4 c1 78 90 6e 08    	kmovw  k5,WORD PTR [r14+0x8]
    1d:	c4 c1 78 90 76 0a    	kmovw  k6,WORD PTR [r14+0xa]
    23:	c4 c1 78 90 7e 0c    	kmovw  k7,WORD PTR [r14+0xc]
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 7.0
  • Core cycles: 7.0
  • Reference cycles: 4.6
  • UOPS_EXECUTED.THREAD: 14.0
  • RETIRE_SLOTS: 21.0
  • UOPS_MITE: 21.0
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.02
  • UOPS_PORT_2: 3.5
  • UOPS_PORT_3: 3.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 7.0
  • UOPS_PORT_6: 0.0
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 7.0

With loop_count=1000 and unroll_count=1

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
     5:	c4 c1 78 90 56 02    	kmovw  k2,WORD PTR [r14+0x2]
     b:	c4 c1 78 90 5e 04    	kmovw  k3,WORD PTR [r14+0x4]
    11:	c4 c1 78 90 66 06    	kmovw  k4,WORD PTR [r14+0x6]
    17:	c4 c1 78 90 6e 08    	kmovw  k5,WORD PTR [r14+0x8]
    1d:	c4 c1 78 90 76 0a    	kmovw  k6,WORD PTR [r14+0xa]
    23:	c4 c1 78 90 7e 0c    	kmovw  k7,WORD PTR [r14+0xc]
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 9.0
  • Core cycles: 7.01
  • Reference cycles: 4.89
  • UOPS_EXECUTED.THREAD: 15.01
  • RETIRE_SLOTS: 22.0
  • UOPS_MITE: 0.02
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 3.5
  • UOPS_PORT_3: 3.51
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 7.0
  • UOPS_PORT_6: 1.0
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.01

With loop_count=1000, unroll_count=1, and padding (long NOPs)

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
     5:	c4 c1 78 90 56 02    	kmovw  k2,WORD PTR [r14+0x2]
     b:	c4 c1 78 90 5e 04    	kmovw  k3,WORD PTR [r14+0x4]
    11:	c4 c1 78 90 66 06    	kmovw  k4,WORD PTR [r14+0x6]
    17:	66 66 66 66 66 66 2e 	data16 data16 data16 data16 data16 cs nop WORD PTR [rax+rax*1+0x0]
    1e:	0f 1f 84 00 00 00 00 
    25:	00 
    26:	c4 c1 78 90 6e 08    	kmovw  k5,WORD PTR [r14+0x8]
    2c:	c4 c1 78 90 76 0a    	kmovw  k6,WORD PTR [r14+0xa]
    32:	c4 c1 78 90 7e 0c    	kmovw  k7,WORD PTR [r14+0xc]
    38:	66 66 66 66 66 66 2e 	data16 data16 data16 data16 data16 cs nop WORD PTR [rax+rax*1+0x0]
    3f:	0f 1f 84 00 00 00 00 
    46:	00 
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 11.0
  • Core cycles: 7.01
  • Reference cycles: 4.81
  • UOPS_EXECUTED.THREAD: 15.01
  • RETIRE_SLOTS: 24.0
  • UOPS_MITE: 0.03
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 3.5
  • UOPS_PORT_3: 3.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 7.0
  • UOPS_PORT_6: 1.0
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.01

With loop_count=100 and unroll_count=10

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
     5:	c4 c1 78 90 56 02    	kmovw  k2,WORD PTR [r14+0x2]
     b:	c4 c1 78 90 5e 04    	kmovw  k3,WORD PTR [r14+0x4]
    11:	c4 c1 78 90 66 06    	kmovw  k4,WORD PTR [r14+0x6]
    17:	c4 c1 78 90 6e 08    	kmovw  k5,WORD PTR [r14+0x8]
    1d:	c4 c1 78 90 76 0a    	kmovw  k6,WORD PTR [r14+0xa]
    23:	c4 c1 78 90 7e 0c    	kmovw  k7,WORD PTR [r14+0xc]
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 7.2
  • Core cycles: 7.01
  • Reference cycles: 4.81
  • UOPS_EXECUTED.THREAD: 14.11
  • RETIRE_SLOTS: 21.1
  • UOPS_MITE: 0.22
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 3.5
  • UOPS_PORT_3: 3.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 7.0
  • UOPS_PORT_6: 0.1
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.07

With loop_count=100, unroll_count=10, and padding (long NOPs)

• Code:

     0:	c4 c1 78 90 0e       	kmovw  k1,WORD PTR [r14]
     5:	c4 c1 78 90 56 02    	kmovw  k2,WORD PTR [r14+0x2]
     b:	c4 c1 78 90 5e 04    	kmovw  k3,WORD PTR [r14+0x4]
    11:	c4 c1 78 90 66 06    	kmovw  k4,WORD PTR [r14+0x6]
    17:	66 66 66 66 66 66 2e 	data16 data16 data16 data16 data16 cs nop WORD PTR [rax+rax*1+0x0]
    1e:	0f 1f 84 00 00 00 00 
    25:	00 
    26:	c4 c1 78 90 6e 08    	kmovw  k5,WORD PTR [r14+0x8]
    2c:	c4 c1 78 90 76 0a    	kmovw  k6,WORD PTR [r14+0xa]
    32:	c4 c1 78 90 7e 0c    	kmovw  k7,WORD PTR [r14+0xc]
    38:	66 66 66 66 66 66 2e 	data16 data16 data16 data16 data16 cs nop WORD PTR [rax+rax*1+0x0]
    3f:	0f 1f 84 00 00 00 00 
    46:	00 
  

• Init:

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

• Show nanoBench command
• Results:
  • Instructions retired: 9.2
  • Core cycles: 7.01
  • Reference cycles: 4.85
  • UOPS_EXECUTED.THREAD: 14.11
  • RETIRE_SLOTS: 23.1
  • UOPS_MITE: 0.23
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 3.5
  • UOPS_PORT_3: 3.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 7.0
  • UOPS_PORT_6: 0.1
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.07

With unroll_count=500 and no inner loop

• Code:

     0:	c4 81 78 90 0c 2e    	kmovw  k1,WORD PTR [r14+r13*1]
  

• 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:
  • Instructions retired: 1.0
  • Core cycles: 1.0
  • Reference cycles: 0.74
  • UOPS_EXECUTED.THREAD: 2.0
  • RETIRE_SLOTS: 3.0
  • UOPS_MITE: 3.0
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 0.5
  • UOPS_PORT_3: 0.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 1.0
  • UOPS_PORT_6: 0.0
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 1.0

With unroll_count=500, no inner loop, and 1 NOP

• Code:

     0:	c4 81 78 90 0c 2e    	kmovw  k1,WORD PTR [r14+r13*1]
     6:	90                   	nop
  

• 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:
  • Instructions retired: 2.0
  • Core cycles: 1.0
  • Reference cycles: 0.58
  • UOPS_EXECUTED.THREAD: 2.0
  • RETIRE_SLOTS: 4.0
  • UOPS_MITE: 4.0
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 0.5
  • UOPS_PORT_3: 0.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 1.0
  • UOPS_PORT_6: 0.0
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 1.0

With unroll_count=500, no inner loop, and 2 NOPs

• Code:

     0:	c4 81 78 90 0c 2e    	kmovw  k1,WORD PTR [r14+r13*1]
     6:	90                   	nop
     7:	90                   	nop
  

• 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:
  • Instructions retired: 3.0
  • Core cycles: 1.25
  • Reference cycles: 0.92
  • UOPS_EXECUTED.THREAD: 2.0
  • RETIRE_SLOTS: 5.0
  • UOPS_MITE: 5.0
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 0.5
  • UOPS_PORT_3: 0.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 1.0
  • UOPS_PORT_6: 0.0
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 1.0

With unroll_count=500, no inner loop, and 3 NOPs

• Code:

     0:	c4 81 78 90 0c 2e    	kmovw  k1,WORD PTR [r14+r13*1]
     6:	90                   	nop
     7:	90                   	nop
     8:	90                   	nop
  

• 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:
  • Instructions retired: 4.0
  • Core cycles: 2.0
  • Reference cycles: 1.44
  • UOPS_EXECUTED.THREAD: 2.0
  • RETIRE_SLOTS: 6.0
  • UOPS_MITE: 6.0
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 0.5
  • UOPS_PORT_3: 0.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 1.0
  • UOPS_PORT_6: 0.0
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 2.0

With loop_count=1000 and unroll_count=10

• Code:

     0:	c4 81 78 90 0c 2e    	kmovw  k1,WORD PTR [r14+r13*1]
  

• 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:
  • Instructions retired: 1.2
  • Core cycles: 1.0
  • Reference cycles: 0.69
  • UOPS_EXECUTED.THREAD: 2.1
  • RETIRE_SLOTS: 3.1
  • UOPS_MITE: 0.0
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 0.5
  • UOPS_PORT_3: 0.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 1.0
  • UOPS_PORT_6: 0.1
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.0

With loop_count=100 and unroll_count=100

• Code:

     0:	c4 81 78 90 0c 2e    	kmovw  k1,WORD PTR [r14+r13*1]
  

• 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:
  • Instructions retired: 1.02
  • Core cycles: 1.0
  • Reference cycles: 0.69
  • UOPS_EXECUTED.THREAD: 2.01
  • RETIRE_SLOTS: 3.01
  • UOPS_MITE: 0.03
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 0.5
  • UOPS_PORT_3: 0.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 1.0
  • UOPS_PORT_6: 0.01
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.01

With unroll_count=200 and no inner loop

• Code:

     0:	c4 81 78 90 0c 2e    	kmovw  k1,WORD PTR [r14+r13*1]
     6:	c4 81 78 90 54 2e 02 	kmovw  k2,WORD PTR [r14+r13*1+0x2]
     d:	c4 81 78 90 5c 2e 04 	kmovw  k3,WORD PTR [r14+r13*1+0x4]
    14:	c4 81 78 90 64 2e 06 	kmovw  k4,WORD PTR [r14+r13*1+0x6]
  

• 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:
  • Instructions retired: 4.0
  • Core cycles: 4.0
  • Reference cycles: 2.99
  • UOPS_EXECUTED.THREAD: 8.0
  • RETIRE_SLOTS: 12.0
  • UOPS_MITE: 12.0
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.01
  • UOPS_PORT_2: 2.0
  • UOPS_PORT_3: 2.0
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 4.0
  • UOPS_PORT_6: 0.0
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 4.0

With loop_count=1000 and unroll_count=2

• Code:

     0:	c4 81 78 90 0c 2e    	kmovw  k1,WORD PTR [r14+r13*1]
     6:	c4 81 78 90 54 2e 02 	kmovw  k2,WORD PTR [r14+r13*1+0x2]
     d:	c4 81 78 90 5c 2e 04 	kmovw  k3,WORD PTR [r14+r13*1+0x4]
    14:	c4 81 78 90 64 2e 06 	kmovw  k4,WORD PTR [r14+r13*1+0x6]
  

• 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:
  • Instructions retired: 5.0
  • Core cycles: 4.01
  • Reference cycles: 2.78
  • UOPS_EXECUTED.THREAD: 8.5
  • RETIRE_SLOTS: 12.5
  • UOPS_MITE: 0.01
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 2.0
  • UOPS_PORT_3: 2.0
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 4.0
  • UOPS_PORT_6: 0.5
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.0

With loop_count=100 and unroll_count=20

• Code:

     0:	c4 81 78 90 0c 2e    	kmovw  k1,WORD PTR [r14+r13*1]
     6:	c4 81 78 90 54 2e 02 	kmovw  k2,WORD PTR [r14+r13*1+0x2]
     d:	c4 81 78 90 5c 2e 04 	kmovw  k3,WORD PTR [r14+r13*1+0x4]
    14:	c4 81 78 90 64 2e 06 	kmovw  k4,WORD PTR [r14+r13*1+0x6]
  

• 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:
  • Instructions retired: 4.1
  • Core cycles: 4.01
  • Reference cycles: 2.78
  • UOPS_EXECUTED.THREAD: 8.05
  • RETIRE_SLOTS: 12.05
  • UOPS_MITE: 0.13
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 2.0
  • UOPS_PORT_3: 2.0
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 4.0
  • UOPS_PORT_6: 0.05
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.04

With unroll_count=100 and no inner loop

• Code:

     0:	c4 81 78 90 0c 2e    	kmovw  k1,WORD PTR [r14+r13*1]
     6:	c4 81 78 90 54 2e 02 	kmovw  k2,WORD PTR [r14+r13*1+0x2]
     d:	c4 81 78 90 5c 2e 04 	kmovw  k3,WORD PTR [r14+r13*1+0x4]
    14:	c4 81 78 90 64 2e 06 	kmovw  k4,WORD PTR [r14+r13*1+0x6]
    1b:	c4 81 78 90 6c 2e 08 	kmovw  k5,WORD PTR [r14+r13*1+0x8]
    22:	c4 81 78 90 74 2e 0a 	kmovw  k6,WORD PTR [r14+r13*1+0xa]
    29:	c4 81 78 90 7c 2e 0c 	kmovw  k7,WORD PTR [r14+r13*1+0xc]
  

• 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:
  • Instructions retired: 7.0
  • Core cycles: 7.0
  • Reference cycles: 4.75
  • UOPS_EXECUTED.THREAD: 14.0
  • RETIRE_SLOTS: 21.0
  • UOPS_MITE: 21.0
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 3.5
  • UOPS_PORT_3: 3.5
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 7.0
  • UOPS_PORT_6: 0.0
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 7.0

With loop_count=1000 and unroll_count=1

• Code:

     0:	c4 81 78 90 0c 2e    	kmovw  k1,WORD PTR [r14+r13*1]
     6:	c4 81 78 90 54 2e 02 	kmovw  k2,WORD PTR [r14+r13*1+0x2]
     d:	c4 81 78 90 5c 2e 04 	kmovw  k3,WORD PTR [r14+r13*1+0x4]
    14:	c4 81 78 90 64 2e 06 	kmovw  k4,WORD PTR [r14+r13*1+0x6]
    1b:	c4 81 78 90 6c 2e 08 	kmovw  k5,WORD PTR [r14+r13*1+0x8]
    22:	c4 81 78 90 74 2e 0a 	kmovw  k6,WORD PTR [r14+r13*1+0xa]
    29:	c4 81 78 90 7c 2e 0c 	kmovw  k7,WORD PTR [r14+r13*1+0xc]
  

• 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:
  • Instructions retired: 9.0
  • Core cycles: 7.01
  • Reference cycles: 4.83
  • UOPS_EXECUTED.THREAD: 15.01
  • RETIRE_SLOTS: 22.0
  • UOPS_MITE: 0.02
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 3.5
  • UOPS_PORT_3: 3.51
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 7.0
  • UOPS_PORT_6: 1.0
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.01

With loop_count=100 and unroll_count=10

• Code:

     0:	c4 81 78 90 0c 2e    	kmovw  k1,WORD PTR [r14+r13*1]
     6:	c4 81 78 90 54 2e 02 	kmovw  k2,WORD PTR [r14+r13*1+0x2]
     d:	c4 81 78 90 5c 2e 04 	kmovw  k3,WORD PTR [r14+r13*1+0x4]
    14:	c4 81 78 90 64 2e 06 	kmovw  k4,WORD PTR [r14+r13*1+0x6]
    1b:	c4 81 78 90 6c 2e 08 	kmovw  k5,WORD PTR [r14+r13*1+0x8]
    22:	c4 81 78 90 74 2e 0a 	kmovw  k6,WORD PTR [r14+r13*1+0xa]
    29:	c4 81 78 90 7c 2e 0c 	kmovw  k7,WORD PTR [r14+r13*1+0xc]
  

• 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:
  • Instructions retired: 7.2
  • Core cycles: 7.01
  • Reference cycles: 4.8
  • UOPS_EXECUTED.THREAD: 14.11
  • RETIRE_SLOTS: 21.11
  • UOPS_MITE: 0.21
  • UOPS_MS: 0.0
  • UOPS_PORT_0: 0.0
  • UOPS_PORT_1: 0.0
  • UOPS_PORT_2: 3.5
  • UOPS_PORT_3: 3.51
  • UOPS_PORT_4: 0.0
  • UOPS_PORT_5: 7.01
  • UOPS_PORT_6: 0.1
  • UOPS_PORT_7: 0.0
  • DIV_CYCLES: 0.0
  • ILD_STALL.LCP: 0.0
  • UOPS_MITE>=1: 0.07