VPCOMPRESSD (M512, K, ZMM) - Port Usage

With blocking instructions for ports {'0', '1'}:

Code:

   0:	c4 e2 e9 98 cb       	vfmadd132pd xmm1,xmm2,xmm3
   5:	c4 e2 e9 98 e3       	vfmadd132pd xmm4,xmm2,xmm3
   a:	c4 e2 e9 98 eb       	vfmadd132pd xmm5,xmm2,xmm3
   f:	c4 e2 e9 98 f3       	vfmadd132pd xmm6,xmm2,xmm3
  14:	c4 e2 e9 98 fb       	vfmadd132pd xmm7,xmm2,xmm3
  19:	c4 62 e9 98 c3       	vfmadd132pd xmm8,xmm2,xmm3
  1e:	c4 62 e9 98 cb       	vfmadd132pd xmm9,xmm2,xmm3
  23:	c4 62 e9 98 d3       	vfmadd132pd xmm10,xmm2,xmm3
  28:	c4 62 e9 98 db       	vfmadd132pd xmm11,xmm2,xmm3
  2d:	c4 62 e9 98 e3       	vfmadd132pd xmm12,xmm2,xmm3
  32:	c4 e2 e9 98 cb       	vfmadd132pd xmm1,xmm2,xmm3
  37:	c4 e2 e9 98 e3       	vfmadd132pd xmm4,xmm2,xmm3
  3c:	c4 e2 e9 98 eb       	vfmadd132pd xmm5,xmm2,xmm3
  41:	c4 e2 e9 98 f3       	vfmadd132pd xmm6,xmm2,xmm3
  46:	c4 e2 e9 98 fb       	vfmadd132pd xmm7,xmm2,xmm3
  4b:	c4 62 e9 98 c3       	vfmadd132pd xmm8,xmm2,xmm3
  50:	62 d2 7d 49 8b 06    	vpcompressd ZMMWORD PTR [r14]{k1},zmm0

Init:

MOV R15, 10000;
L: VADDPS ZMM0, ZMM1, ZMM1;
VADDPS ZMM0, ZMM1, ZMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
VMOVUPD XMM1, [R14];
VMOVUPD XMM2, [R14];
VMOVUPD XMM3, [R14];
VMOVUPD XMM4, [R14];
VMOVUPD XMM5, [R14];
VMOVUPD XMM6, [R14];
VMOVUPD XMM7, [R14];
VMOVUPD XMM8, [R14];
VMOVUPD XMM9, [R14];
VMOVUPD XMM10, [R14];
VMOVUPD XMM11, [R14];
VMOVUPD XMM12, [R14];
VXORPS ZMM0, ZMM0, ZMM0

Show nanoBench command
Results:
- MPERF: 6.4
- APERF: 9.02
- FpuPipeAssignment.Total0: 7.92
- FpuPipeAssignment.Total1: 8.25
- FpuPipeAssignment.Total2: 1.92
- FpuPipeAssignment.Total3: 1.91
- FpuPipeAssignment.Total4: 0.5
- FpuPipeAssignment.Total5: 0.5

With blocking instructions for ports {'0', '3'}:

Code:

   0:	0f ec c1             	paddsb mm0,mm1
   3:	0f ec d1             	paddsb mm2,mm1
   6:	0f ec d9             	paddsb mm3,mm1
   9:	0f ec e1             	paddsb mm4,mm1
   c:	0f ec e9             	paddsb mm5,mm1
   f:	0f ec f1             	paddsb mm6,mm1
  12:	0f ec f9             	paddsb mm7,mm1
  15:	0f ec c1             	paddsb mm0,mm1
  18:	0f ec d1             	paddsb mm2,mm1
  1b:	0f ec d9             	paddsb mm3,mm1
  1e:	0f ec e1             	paddsb mm4,mm1
  21:	0f ec e9             	paddsb mm5,mm1
  24:	0f ec f1             	paddsb mm6,mm1
  27:	0f ec f9             	paddsb mm7,mm1
  2a:	0f ec c1             	paddsb mm0,mm1
  2d:	0f ec d1             	paddsb mm2,mm1
  30:	62 d2 7d 49 8b 06    	vpcompressd ZMMWORD PTR [r14]{k1},zmm0

Init:

MOV R15, 10000;
L: VADDPS ZMM0, ZMM1, ZMM1;
VADDPS ZMM0, ZMM1, ZMM1;
DEC R15;
JNZ L;
PXOR MM0, MM0;
PXOR MM1, MM1;
PXOR MM2, MM2;
PXOR MM3, MM3;
PXOR MM4, MM4;
PXOR MM5, MM5;
PXOR MM6, MM6;
PXOR MM7, MM7;
VZEROALL;
VXORPS ZMM0, ZMM0, ZMM0

Show nanoBench command
Results:
- MPERF: 6.27
- APERF: 9.16
- FpuPipeAssignment.Total0: 9.0
- FpuPipeAssignment.Total1: 1.0
- FpuPipeAssignment.Total2: 1.0
- FpuPipeAssignment.Total3: 9.0
- FpuPipeAssignment.Total4: 0.5
- FpuPipeAssignment.Total5: 0.5

⇨ 2 μops that can only use ports {'0', '3'}

With blocking instructions for ports {'1', '2'}:

Code:

   0:	0f 6b c1             	packssdw mm0,mm1
   3:	0f 6b d1             	packssdw mm2,mm1
   6:	0f 6b d9             	packssdw mm3,mm1
   9:	0f 6b e1             	packssdw mm4,mm1
   c:	0f 6b e9             	packssdw mm5,mm1
   f:	0f 6b f1             	packssdw mm6,mm1
  12:	0f 6b f9             	packssdw mm7,mm1
  15:	0f 6b c1             	packssdw mm0,mm1
  18:	0f 6b d1             	packssdw mm2,mm1
  1b:	0f 6b d9             	packssdw mm3,mm1
  1e:	0f 6b e1             	packssdw mm4,mm1
  21:	0f 6b e9             	packssdw mm5,mm1
  24:	0f 6b f1             	packssdw mm6,mm1
  27:	0f 6b f9             	packssdw mm7,mm1
  2a:	0f 6b c1             	packssdw mm0,mm1
  2d:	0f 6b d1             	packssdw mm2,mm1
  30:	62 d2 7d 49 8b 06    	vpcompressd ZMMWORD PTR [r14]{k1},zmm0

Init:

MOV R15, 10000;
L: VADDPS ZMM0, ZMM1, ZMM1;
VADDPS ZMM0, ZMM1, ZMM1;
DEC R15;
JNZ L;
PXOR MM0, MM0;
PXOR MM1, MM1;
PXOR MM2, MM2;
PXOR MM3, MM3;
PXOR MM4, MM4;
PXOR MM5, MM5;
PXOR MM6, MM6;
PXOR MM7, MM7;
VZEROALL;
VXORPS ZMM0, ZMM0, ZMM0

Show nanoBench command
Results:
- MPERF: 6.4
- APERF: 9.26
- FpuPipeAssignment.Total0: 1.0
- FpuPipeAssignment.Total1: 9.0
- FpuPipeAssignment.Total2: 9.0
- FpuPipeAssignment.Total3: 1.0
- FpuPipeAssignment.Total4: 0.5
- FpuPipeAssignment.Total5: 0.5

⇨ 2 μops that can only use ports {'1', '2'}

With blocking instructions for ports {'2', '3'}:

Code:

   0:	0f 72 f0 02          	pslld  mm0,0x2
   4:	0f 72 f1 02          	pslld  mm1,0x2
   8:	0f 72 f2 02          	pslld  mm2,0x2
   c:	0f 72 f3 02          	pslld  mm3,0x2
  10:	0f 72 f4 02          	pslld  mm4,0x2
  14:	0f 72 f5 02          	pslld  mm5,0x2
  18:	0f 72 f6 02          	pslld  mm6,0x2
  1c:	0f 72 f7 02          	pslld  mm7,0x2
  20:	0f 72 f0 02          	pslld  mm0,0x2
  24:	0f 72 f1 02          	pslld  mm1,0x2
  28:	0f 72 f2 02          	pslld  mm2,0x2
  2c:	0f 72 f3 02          	pslld  mm3,0x2
  30:	0f 72 f4 02          	pslld  mm4,0x2
  34:	0f 72 f5 02          	pslld  mm5,0x2
  38:	0f 72 f6 02          	pslld  mm6,0x2
  3c:	0f 72 f7 02          	pslld  mm7,0x2
  40:	62 d2 7d 49 8b 06    	vpcompressd ZMMWORD PTR [r14]{k1},zmm0

Init:

MOV R15, 10000;
L: VADDPS ZMM0, ZMM1, ZMM1;
VADDPS ZMM0, ZMM1, ZMM1;
DEC R15;
JNZ L;
PXOR MM0, MM0;
PXOR MM1, MM1;
PXOR MM2, MM2;
PXOR MM3, MM3;
PXOR MM4, MM4;
PXOR MM5, MM5;
PXOR MM6, MM6;
PXOR MM7, MM7;
VZEROALL;
VXORPS ZMM0, ZMM0, ZMM0

Show nanoBench command
Results:
- MPERF: 5.64
- APERF: 8.22
- FpuPipeAssignment.Total0: 1.93
- FpuPipeAssignment.Total1: 1.88
- FpuPipeAssignment.Total2: 8.13
- FpuPipeAssignment.Total3: 8.07
- FpuPipeAssignment.Total4: 0.5
- FpuPipeAssignment.Total5: 0.5

With blocking instructions for ports {'4', '5'}:

Code:

   0:	41 0f 7e c0          	movd   r8d,mm0
   4:	41 0f 7e c1          	movd   r9d,mm0
   8:	41 0f 7e c2          	movd   r10d,mm0
   c:	41 0f 7e c3          	movd   r11d,mm0
  10:	41 0f 7e c4          	movd   r12d,mm0
  14:	0f 7e c3             	movd   ebx,mm0
  17:	0f 7e c1             	movd   ecx,mm0
  1a:	0f 7e c2             	movd   edx,mm0
  1d:	0f 7e c0             	movd   eax,mm0
  20:	41 0f 7e c0          	movd   r8d,mm0
  24:	41 0f 7e c1          	movd   r9d,mm0
  28:	41 0f 7e c2          	movd   r10d,mm0
  2c:	41 0f 7e c3          	movd   r11d,mm0
  30:	41 0f 7e c4          	movd   r12d,mm0
  34:	0f 7e c3             	movd   ebx,mm0
  37:	0f 7e c1             	movd   ecx,mm0
  3a:	62 d2 7d 49 8b 06    	vpcompressd ZMMWORD PTR [r14]{k1},zmm0

Init:

MOV R15, 10000;
L: VADDPS ZMM0, ZMM1, ZMM1;
VADDPS ZMM0, ZMM1, ZMM1;
DEC R15;
JNZ L;
PXOR MM0, MM0;
VZEROALL;
VXORPS ZMM0, ZMM0, ZMM0

Show nanoBench command
Results:
- MPERF: 6.46
- APERF: 9.14
- FpuPipeAssignment.Total0: 1.0
- FpuPipeAssignment.Total1: 1.01
- FpuPipeAssignment.Total2: 1.0
- FpuPipeAssignment.Total3: 1.0
- FpuPipeAssignment.Total4: 8.5
- FpuPipeAssignment.Total5: 8.5

⇨ One μop that can only use ports {'4', '5'}

With an indexed addressing mode

With blocking instructions for ports {'0', '1'}:

Code:

   0:	c4 e2 e9 98 cb       	vfmadd132pd xmm1,xmm2,xmm3
   5:	c4 e2 e9 98 e3       	vfmadd132pd xmm4,xmm2,xmm3
   a:	c4 e2 e9 98 eb       	vfmadd132pd xmm5,xmm2,xmm3
   f:	c4 e2 e9 98 f3       	vfmadd132pd xmm6,xmm2,xmm3
  14:	c4 e2 e9 98 fb       	vfmadd132pd xmm7,xmm2,xmm3
  19:	c4 62 e9 98 c3       	vfmadd132pd xmm8,xmm2,xmm3
  1e:	c4 62 e9 98 cb       	vfmadd132pd xmm9,xmm2,xmm3
  23:	c4 62 e9 98 d3       	vfmadd132pd xmm10,xmm2,xmm3
  28:	c4 62 e9 98 db       	vfmadd132pd xmm11,xmm2,xmm3
  2d:	c4 62 e9 98 e3       	vfmadd132pd xmm12,xmm2,xmm3
  32:	c4 e2 e9 98 cb       	vfmadd132pd xmm1,xmm2,xmm3
  37:	c4 e2 e9 98 e3       	vfmadd132pd xmm4,xmm2,xmm3
  3c:	c4 e2 e9 98 eb       	vfmadd132pd xmm5,xmm2,xmm3
  41:	c4 e2 e9 98 f3       	vfmadd132pd xmm6,xmm2,xmm3
  46:	c4 e2 e9 98 fb       	vfmadd132pd xmm7,xmm2,xmm3
  4b:	c4 62 e9 98 c3       	vfmadd132pd xmm8,xmm2,xmm3
  50:	62 92 7d 49 8b 04 2e 	vpcompressd ZMMWORD PTR [r14+r13*1]{k1},zmm0

Init:

MOV R15, 10000;
L: VADDPS ZMM0, ZMM1, ZMM1;
VADDPS ZMM0, ZMM1, ZMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
VMOVUPD XMM1, [R14];
VMOVUPD XMM2, [R14];
VMOVUPD XMM3, [R14];
VMOVUPD XMM4, [R14];
VMOVUPD XMM5, [R14];
VMOVUPD XMM6, [R14];
VMOVUPD XMM7, [R14];
VMOVUPD XMM8, [R14];
VMOVUPD XMM9, [R14];
VMOVUPD XMM10, [R14];
VMOVUPD XMM11, [R14];
VMOVUPD XMM12, [R14];
VXORPS ZMM0, ZMM0, ZMM0;
XOR R13, R13

Show nanoBench command
Results:
- MPERF: 6.08
- APERF: 9.02
- FpuPipeAssignment.Total0: 7.93
- FpuPipeAssignment.Total1: 8.24
- FpuPipeAssignment.Total2: 1.93
- FpuPipeAssignment.Total3: 1.91
- FpuPipeAssignment.Total4: 0.5
- FpuPipeAssignment.Total5: 0.5

With blocking instructions for ports {'0', '3'}:

Code:

   0:	0f ec c1             	paddsb mm0,mm1
   3:	0f ec d1             	paddsb mm2,mm1
   6:	0f ec d9             	paddsb mm3,mm1
   9:	0f ec e1             	paddsb mm4,mm1
   c:	0f ec e9             	paddsb mm5,mm1
   f:	0f ec f1             	paddsb mm6,mm1
  12:	0f ec f9             	paddsb mm7,mm1
  15:	0f ec c1             	paddsb mm0,mm1
  18:	0f ec d1             	paddsb mm2,mm1
  1b:	0f ec d9             	paddsb mm3,mm1
  1e:	0f ec e1             	paddsb mm4,mm1
  21:	0f ec e9             	paddsb mm5,mm1
  24:	0f ec f1             	paddsb mm6,mm1
  27:	0f ec f9             	paddsb mm7,mm1
  2a:	0f ec c1             	paddsb mm0,mm1
  2d:	0f ec d1             	paddsb mm2,mm1
  30:	62 92 7d 49 8b 04 2e 	vpcompressd ZMMWORD PTR [r14+r13*1]{k1},zmm0

Init:

MOV R15, 10000;
L: VADDPS ZMM0, ZMM1, ZMM1;
VADDPS ZMM0, ZMM1, ZMM1;
DEC R15;
JNZ L;
PXOR MM0, MM0;
PXOR MM1, MM1;
PXOR MM2, MM2;
PXOR MM3, MM3;
PXOR MM4, MM4;
PXOR MM5, MM5;
PXOR MM6, MM6;
PXOR MM7, MM7;
VZEROALL;
VXORPS ZMM0, ZMM0, ZMM0;
XOR R13, R13

Show nanoBench command
Results:
- MPERF: 6.4
- APERF: 9.17
- FpuPipeAssignment.Total0: 9.0
- FpuPipeAssignment.Total1: 1.0
- FpuPipeAssignment.Total2: 1.0
- FpuPipeAssignment.Total3: 9.0
- FpuPipeAssignment.Total4: 0.5
- FpuPipeAssignment.Total5: 0.5

⇨ 2 μops that can only use ports {'0', '3'}

With blocking instructions for ports {'1', '2'}:

Code:

   0:	0f 6b c1             	packssdw mm0,mm1
   3:	0f 6b d1             	packssdw mm2,mm1
   6:	0f 6b d9             	packssdw mm3,mm1
   9:	0f 6b e1             	packssdw mm4,mm1
   c:	0f 6b e9             	packssdw mm5,mm1
   f:	0f 6b f1             	packssdw mm6,mm1
  12:	0f 6b f9             	packssdw mm7,mm1
  15:	0f 6b c1             	packssdw mm0,mm1
  18:	0f 6b d1             	packssdw mm2,mm1
  1b:	0f 6b d9             	packssdw mm3,mm1
  1e:	0f 6b e1             	packssdw mm4,mm1
  21:	0f 6b e9             	packssdw mm5,mm1
  24:	0f 6b f1             	packssdw mm6,mm1
  27:	0f 6b f9             	packssdw mm7,mm1
  2a:	0f 6b c1             	packssdw mm0,mm1
  2d:	0f 6b d1             	packssdw mm2,mm1
  30:	62 92 7d 49 8b 04 2e 	vpcompressd ZMMWORD PTR [r14+r13*1]{k1},zmm0

Init:

MOV R15, 10000;
L: VADDPS ZMM0, ZMM1, ZMM1;
VADDPS ZMM0, ZMM1, ZMM1;
DEC R15;
JNZ L;
PXOR MM0, MM0;
PXOR MM1, MM1;
PXOR MM2, MM2;
PXOR MM3, MM3;
PXOR MM4, MM4;
PXOR MM5, MM5;
PXOR MM6, MM6;
PXOR MM7, MM7;
VZEROALL;
VXORPS ZMM0, ZMM0, ZMM0;
XOR R13, R13

Show nanoBench command
Results:
- MPERF: 6.46
- APERF: 9.28
- FpuPipeAssignment.Total0: 1.0
- FpuPipeAssignment.Total1: 9.0
- FpuPipeAssignment.Total2: 9.0
- FpuPipeAssignment.Total3: 1.0
- FpuPipeAssignment.Total4: 0.5
- FpuPipeAssignment.Total5: 0.5

⇨ 2 μops that can only use ports {'1', '2'}

With blocking instructions for ports {'2', '3'}:

Code:

   0:	0f 72 f0 02          	pslld  mm0,0x2
   4:	0f 72 f1 02          	pslld  mm1,0x2
   8:	0f 72 f2 02          	pslld  mm2,0x2
   c:	0f 72 f3 02          	pslld  mm3,0x2
  10:	0f 72 f4 02          	pslld  mm4,0x2
  14:	0f 72 f5 02          	pslld  mm5,0x2
  18:	0f 72 f6 02          	pslld  mm6,0x2
  1c:	0f 72 f7 02          	pslld  mm7,0x2
  20:	0f 72 f0 02          	pslld  mm0,0x2
  24:	0f 72 f1 02          	pslld  mm1,0x2
  28:	0f 72 f2 02          	pslld  mm2,0x2
  2c:	0f 72 f3 02          	pslld  mm3,0x2
  30:	0f 72 f4 02          	pslld  mm4,0x2
  34:	0f 72 f5 02          	pslld  mm5,0x2
  38:	0f 72 f6 02          	pslld  mm6,0x2
  3c:	0f 72 f7 02          	pslld  mm7,0x2
  40:	62 92 7d 49 8b 04 2e 	vpcompressd ZMMWORD PTR [r14+r13*1]{k1},zmm0

Init:

MOV R15, 10000;
L: VADDPS ZMM0, ZMM1, ZMM1;
VADDPS ZMM0, ZMM1, ZMM1;
DEC R15;
JNZ L;
PXOR MM0, MM0;
PXOR MM1, MM1;
PXOR MM2, MM2;
PXOR MM3, MM3;
PXOR MM4, MM4;
PXOR MM5, MM5;
PXOR MM6, MM6;
PXOR MM7, MM7;
VZEROALL;
VXORPS ZMM0, ZMM0, ZMM0;
XOR R13, R13

Show nanoBench command
Results:
- MPERF: 5.7
- APERF: 8.22
- FpuPipeAssignment.Total0: 1.93
- FpuPipeAssignment.Total1: 1.88
- FpuPipeAssignment.Total2: 8.13
- FpuPipeAssignment.Total3: 8.07
- FpuPipeAssignment.Total4: 0.5
- FpuPipeAssignment.Total5: 0.5

With blocking instructions for ports {'4', '5'}:

Code:

   0:	41 0f 7e c0          	movd   r8d,mm0
   4:	41 0f 7e c1          	movd   r9d,mm0
   8:	41 0f 7e c2          	movd   r10d,mm0
   c:	41 0f 7e c3          	movd   r11d,mm0
  10:	41 0f 7e c4          	movd   r12d,mm0
  14:	0f 7e c3             	movd   ebx,mm0
  17:	0f 7e c1             	movd   ecx,mm0
  1a:	0f 7e c2             	movd   edx,mm0
  1d:	0f 7e c0             	movd   eax,mm0
  20:	41 0f 7e c0          	movd   r8d,mm0
  24:	41 0f 7e c1          	movd   r9d,mm0
  28:	41 0f 7e c2          	movd   r10d,mm0
  2c:	41 0f 7e c3          	movd   r11d,mm0
  30:	41 0f 7e c4          	movd   r12d,mm0
  34:	0f 7e c3             	movd   ebx,mm0
  37:	0f 7e c1             	movd   ecx,mm0
  3a:	62 92 7d 49 8b 04 2e 	vpcompressd ZMMWORD PTR [r14+r13*1]{k1},zmm0

Init:

MOV R15, 10000;
L: VADDPS ZMM0, ZMM1, ZMM1;
VADDPS ZMM0, ZMM1, ZMM1;
DEC R15;
JNZ L;
PXOR MM0, MM0;
VZEROALL;
VXORPS ZMM0, ZMM0, ZMM0;
XOR R13, R13

Show nanoBench command
Results:
- MPERF: 6.46
- APERF: 9.14
- FpuPipeAssignment.Total0: 1.0
- FpuPipeAssignment.Total1: 1.0
- FpuPipeAssignment.Total2: 1.0
- FpuPipeAssignment.Total3: 1.0
- FpuPipeAssignment.Total4: 8.5
- FpuPipeAssignment.Total5: 8.5

⇨ One μop that can only use ports {'4', '5'}