VGATHERDPS (YMM, VSIB

Experiment 1

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instruction: VPCMPEQD YMM1, YMM1, YMM1

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 7.83
- APERF: 9.65
- UOPS: 40.0

Experiment 2

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: 11

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
   a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
   e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  12:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  16:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  22:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  26:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  32:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 17.08
- APERF: 21.0
- UOPS: 51.0

Experiment 3

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: 11

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
   a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
   e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  12:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  16:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  22:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  26:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  32:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 15.54
- APERF: 19.0
- UOPS: 51.0

Experiment 4 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instruction: VPCMPEQD YMM1, YMM1, YMM1

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM0, [R14+YMM14], YMM1;
VGATHERDPS YMM1, [R14+YMM14], YMM0

Show nanoBench command
Results:
- MPERF: 7.77
- APERF: 9.62
- UOPS: 40.0

Experiment 5 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: 11

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
   a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
   e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  12:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  16:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  22:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  26:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  32:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM0, [R14+YMM14], YMM1;
VGATHERDPS YMM1, [R14+YMM14], YMM0

Show nanoBench command
Results:
- MPERF: 17.14
- APERF: 21.0
- UOPS: 51.0

Experiment 6 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: 11

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
   a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
   e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  12:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  16:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  22:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  26:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  32:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM0, [R14+YMM14], YMM1;
VGATHERDPS YMM1, [R14+YMM14], YMM0

Show nanoBench command
Results:
- MPERF: 15.54
- APERF: 19.0
- UOPS: 51.0

Latency operand 1 → 3: 0

Experiment 1

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VANDPD YMM0, YMM1, YMM1;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: 11

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 f5 54 c1          	vandpd ymm0,ymm1,ymm1
   a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
   e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  12:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  16:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  22:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  26:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  32:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 9.25
- APERF: 11.46
- UOPS: 51.0

Experiment 2

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VPAND YMM0, YMM1, YMM1;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: 11

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 f5 db c1          	vpand  ymm0,ymm1,ymm1
   a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
   e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  12:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  16:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  22:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  26:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  32:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 9.62
- APERF: 11.62
- UOPS: 51.0

Experiment 3 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VANDPD YMM0, YMM1, YMM1;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: 11

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 f5 54 c1          	vandpd ymm0,ymm1,ymm1
   a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
   e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  12:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  16:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  22:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  26:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  32:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM0, [R14+YMM14], YMM1;
VGATHERDPS YMM1, [R14+YMM14], YMM0

Show nanoBench command
Results:
- MPERF: 9.56
- APERF: 11.56
- UOPS: 51.0

Experiment 4 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VPAND YMM0, YMM1, YMM1;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: 11

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 f5 db c1          	vpand  ymm0,ymm1,ymm1
   a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
   e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  12:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  16:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  22:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  26:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  32:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM0, [R14+YMM14], YMM1;
VGATHERDPS YMM1, [R14+YMM14], YMM0

Show nanoBench command
Results:
- MPERF: 9.25
- APERF: 11.38
- UOPS: 51.0

Latency operand 2 → 1 (address, base register): ≤20

Experiment 1

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VMOVQ R12, XMM0;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: ≥11

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c4 c1 f9 7e c4       	vmovq  r12,xmm0
   b:	4d 31 e6             	xor    r14,r12
   e:	4d 31 e6             	xor    r14,r12
  11:	4d 31 e6             	xor    r14,r12
  14:	4d 31 e6             	xor    r14,r12
  17:	4d 31 e6             	xor    r14,r12
  1a:	4d 31 e6             	xor    r14,r12
  1d:	4d 31 e6             	xor    r14,r12
  20:	4d 31 e6             	xor    r14,r12
  23:	4d 31 e6             	xor    r14,r12
  26:	4d 31 e6             	xor    r14,r12
  29:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 25.53
- APERF: 31.0
- UOPS: 51.0

Experiment 2 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VMOVQ R12, XMM0;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: ≥11

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c4 c1 f9 7e c4       	vmovq  r12,xmm0
   b:	4d 31 e6             	xor    r14,r12
   e:	4d 31 e6             	xor    r14,r12
  11:	4d 31 e6             	xor    r14,r12
  14:	4d 31 e6             	xor    r14,r12
  17:	4d 31 e6             	xor    r14,r12
  1a:	4d 31 e6             	xor    r14,r12
  1d:	4d 31 e6             	xor    r14,r12
  20:	4d 31 e6             	xor    r14,r12
  23:	4d 31 e6             	xor    r14,r12
  26:	4d 31 e6             	xor    r14,r12
  29:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM0, [R14+YMM14], YMM1;
VGATHERDPS YMM1, [R14+YMM14], YMM0

Show nanoBench command
Results:
- MPERF: 25.41
- APERF: 31.0
- UOPS: 51.0

Latency operand 2 → 1 (address, index register): ≤28

Experiment 1

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: ≥10

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
   a:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
   e:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  12:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  16:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  1a:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  1e:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  22:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  26:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  2a:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  2e:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 31.51
- APERF: 38.65
- UOPS: 50.0

Experiment 2 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: ≥10

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
   a:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
   e:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  12:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  16:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  1a:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  1e:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  22:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  26:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  2a:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  2e:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM0, [R14+YMM14], YMM1;
VGATHERDPS YMM1, [R14+YMM14], YMM0

Show nanoBench command
Results:
- MPERF: 31.51
- APERF: 38.75
- UOPS: 50.0

Latency operand 2 → 1 (memory): ≤16

Experiment 1

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VANDPD XMM13, XMM0, XMM0;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VMOVSS [R14], XMM13;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: ≥12

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 79 54 e8          	vandpd xmm13,xmm0,xmm0
   a:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
   f:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  14:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  19:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  1e:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  23:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  28:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  2d:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  32:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  37:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  3c:	c4 41 7a 11 2e       	vmovss DWORD PTR [r14],xmm13
  41:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 23.68
- APERF: 29.0
- UOPS: 52.0

Experiment 2

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VPAND XMM13, XMM0, XMM0;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VMOVD [R14], XMM13;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: ≥12

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 79 db e8          	vpand  xmm13,xmm0,xmm0
   a:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
   f:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  14:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  19:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  1e:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  23:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  28:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  2d:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  32:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  37:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  3c:	c4 41 79 7e 2e       	vmovd  DWORD PTR [r14],xmm13
  41:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 22.94
- APERF: 28.01
- UOPS: 52.0

Experiment 3 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VANDPD XMM13, XMM0, XMM0;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VMOVSS [R14], XMM13;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: ≥12

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 79 54 e8          	vandpd xmm13,xmm0,xmm0
   a:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
   f:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  14:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  19:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  1e:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  23:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  28:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  2d:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  32:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  37:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  3c:	c4 41 7a 11 2e       	vmovss DWORD PTR [r14],xmm13
  41:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM0, [R14+YMM14], YMM1;
VGATHERDPS YMM1, [R14+YMM14], YMM0

Show nanoBench command
Results:
- MPERF: 23.68
- APERF: 29.0
- UOPS: 52.0

Experiment 4 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM0, [R14+YMM14], YMM1
Chain instructions: VPAND XMM13, XMM0, XMM0;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VMOVD [R14], XMM13;VPCMPEQD YMM1, YMM1, YMM1
Chain latency: ≥12

Code:

   0:	c4 82 75 92 04 36    	vgatherdps ymm0,DWORD PTR [r14+ymm14*1],ymm1
   6:	c5 79 db e8          	vpand  xmm13,xmm0,xmm0
   a:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
   f:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  14:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  19:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  1e:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  23:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  28:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  2d:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  32:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  37:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  3c:	c4 41 79 7e 2e       	vmovd  DWORD PTR [r14],xmm13
  41:	c5 f5 76 c9          	vpcmpeqd ymm1,ymm1,ymm1

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM0, [R14];
VXORPS YMM1, YMM1, YMM1;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM0, [R14+YMM14], YMM1;
VGATHERDPS YMM1, [R14+YMM14], YMM0

Show nanoBench command
Results:
- MPERF: 22.76
- APERF: 28.02
- UOPS: 52.0

Latency operand 2 → 3 (address, base register): ≤6

Experiment 1

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VMOVQ R12, XMM0;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;VPCMPEQD YMM0, YMM0, YMM0
Chain latency: ≥11

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c4 c1 f9 7e c4       	vmovq  r12,xmm0
   b:	4d 31 e6             	xor    r14,r12
   e:	4d 31 e6             	xor    r14,r12
  11:	4d 31 e6             	xor    r14,r12
  14:	4d 31 e6             	xor    r14,r12
  17:	4d 31 e6             	xor    r14,r12
  1a:	4d 31 e6             	xor    r14,r12
  1d:	4d 31 e6             	xor    r14,r12
  20:	4d 31 e6             	xor    r14,r12
  23:	4d 31 e6             	xor    r14,r12
  26:	4d 31 e6             	xor    r14,r12
  29:	c5 fd 76 c0          	vpcmpeqd ymm0,ymm0,ymm0

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 14.43
- APERF: 17.41
- UOPS: 51.0

Experiment 2 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VMOVQ R12, XMM0;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;XOR R14, R12;VPCMPEQD YMM0, YMM0, YMM0
Chain latency: ≥11

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c4 c1 f9 7e c4       	vmovq  r12,xmm0
   b:	4d 31 e6             	xor    r14,r12
   e:	4d 31 e6             	xor    r14,r12
  11:	4d 31 e6             	xor    r14,r12
  14:	4d 31 e6             	xor    r14,r12
  17:	4d 31 e6             	xor    r14,r12
  1a:	4d 31 e6             	xor    r14,r12
  1d:	4d 31 e6             	xor    r14,r12
  20:	4d 31 e6             	xor    r14,r12
  23:	4d 31 e6             	xor    r14,r12
  26:	4d 31 e6             	xor    r14,r12
  29:	c5 fd 76 c0          	vpcmpeqd ymm0,ymm0,ymm0

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM1, [R14+YMM14], YMM0;
VGATHERDPS YMM0, [R14+YMM14], YMM1

Show nanoBench command
Results:
- MPERF: 14.43
- APERF: 17.46
- UOPS: 51.0

Latency operand 2 → 3 (address, index register): ≤1

Experiment 1

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VPCMPEQD YMM0, YMM0, YMM0
Chain latency: ≥10

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
   a:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
   e:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  12:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  16:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  1a:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  1e:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  22:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  26:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  2a:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  2e:	c5 fd 76 c0          	vpcmpeqd ymm0,ymm0,ymm0

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 9.25
- APERF: 11.28
- UOPS: 50.0

Experiment 2 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VANDPD YMM14, YMM14, YMM0;VPCMPEQD YMM0, YMM0, YMM0
Chain latency: ≥10

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
   a:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
   e:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  12:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  16:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  1a:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  1e:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  22:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  26:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  2a:	c5 0d 54 f0          	vandpd ymm14,ymm14,ymm0
  2e:	c5 fd 76 c0          	vpcmpeqd ymm0,ymm0,ymm0

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM1, [R14+YMM14], YMM0;
VGATHERDPS YMM0, [R14+YMM14], YMM1

Show nanoBench command
Results:
- MPERF: 9.25
- APERF: 11.28
- UOPS: 50.0

Latency operand 2 → 3 (memory): ≤0

Experiment 1

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VANDPD XMM13, XMM0, XMM0;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VMOVSS [R14], XMM13;VPCMPEQD YMM0, YMM0, YMM0
Chain latency: ≥12

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c5 79 54 e8          	vandpd xmm13,xmm0,xmm0
   a:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
   f:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  14:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  19:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  1e:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  23:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  28:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  2d:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  32:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  37:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  3c:	c4 41 7a 11 2e       	vmovss DWORD PTR [r14],xmm13
  41:	c5 fd 76 c0          	vpcmpeqd ymm0,ymm0,ymm0

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 8.2
- APERF: 10.0
- UOPS: 52.0

Experiment 2

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VPAND XMM13, XMM0, XMM0;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VMOVD [R14], XMM13;VPCMPEQD YMM0, YMM0, YMM0
Chain latency: ≥12

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c5 79 db e8          	vpand  xmm13,xmm0,xmm0
   a:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
   f:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  14:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  19:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  1e:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  23:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  28:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  2d:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  32:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  37:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  3c:	c4 41 79 7e 2e       	vmovd  DWORD PTR [r14],xmm13
  41:	c5 fd 76 c0          	vpcmpeqd ymm0,ymm0,ymm0

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 8.14
- APERF: 10.0
- UOPS: 52.0

Experiment 3 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VANDPD XMM13, XMM0, XMM0;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VANDPD XMM13, XMM13, XMM13;VMOVSS [R14], XMM13;VPCMPEQD YMM0, YMM0, YMM0
Chain latency: ≥12

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c5 79 54 e8          	vandpd xmm13,xmm0,xmm0
   a:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
   f:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  14:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  19:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  1e:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  23:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  28:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  2d:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  32:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  37:	c4 41 11 54 ed       	vandpd xmm13,xmm13,xmm13
  3c:	c4 41 7a 11 2e       	vmovss DWORD PTR [r14],xmm13
  41:	c5 fd 76 c0          	vpcmpeqd ymm0,ymm0,ymm0

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM1, [R14+YMM14], YMM0;
VGATHERDPS YMM0, [R14+YMM14], YMM1

Show nanoBench command
Results:
- MPERF: 8.08
- APERF: 10.0
- UOPS: 52.0

Experiment 4 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VPAND XMM13, XMM0, XMM0;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VPAND XMM13, XMM13, XMM13;VMOVD [R14], XMM13;VPCMPEQD YMM0, YMM0, YMM0
Chain latency: ≥12

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c5 79 db e8          	vpand  xmm13,xmm0,xmm0
   a:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
   f:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  14:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  19:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  1e:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  23:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  28:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  2d:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  32:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  37:	c4 41 11 db ed       	vpand  xmm13,xmm13,xmm13
  3c:	c4 41 79 7e 2e       	vmovd  DWORD PTR [r14],xmm13
  41:	c5 fd 76 c0          	vpcmpeqd ymm0,ymm0,ymm0

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM1, [R14+YMM14], YMM0;
VGATHERDPS YMM0, [R14+YMM14], YMM1

Show nanoBench command
Results:
- MPERF: 8.2
- APERF: 10.0
- UOPS: 52.0

Latency operand 3 → 1: 19

Experiment 1

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VANDPD YMM0, YMM1, YMM1;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VPCMPEQD YMM13, YMM13, YMM13;VPOR YMM0, YMM0, YMM13
Chain latency: 12

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c5 f5 54 c1          	vandpd ymm0,ymm1,ymm1
   a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
   e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  12:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  16:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  22:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  26:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  32:	c4 41 15 76 ed       	vpcmpeqd ymm13,ymm13,ymm13
  37:	c4 c1 7d eb c5       	vpor   ymm0,ymm0,ymm13

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 27.2
- APERF: 33.02
- UOPS: 52.0

Experiment 2

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VPAND YMM0, YMM1, YMM1;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPCMPEQD YMM13, YMM13, YMM13;VPOR YMM0, YMM0, YMM13
Chain latency: 12

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c5 f5 db c1          	vpand  ymm0,ymm1,ymm1
   a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
   e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  12:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  16:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  22:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  26:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  32:	c4 41 15 76 ed       	vpcmpeqd ymm13,ymm13,ymm13
  37:	c4 c1 7d eb c5       	vpor   ymm0,ymm0,ymm13

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 25.16
- APERF: 31.0
- UOPS: 52.0

Experiment 3 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VANDPD YMM0, YMM1, YMM1;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VPCMPEQD YMM13, YMM13, YMM13;VPOR YMM0, YMM0, YMM13
Chain latency: 12

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c5 f5 54 c1          	vandpd ymm0,ymm1,ymm1
   a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
   e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  12:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  16:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  22:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  26:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  32:	c4 41 15 76 ed       	vpcmpeqd ymm13,ymm13,ymm13
  37:	c4 c1 7d eb c5       	vpor   ymm0,ymm0,ymm13

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM1, [R14+YMM14], YMM0;
VGATHERDPS YMM0, [R14+YMM14], YMM1

Show nanoBench command
Results:
- MPERF: 27.38
- APERF: 33.1
- UOPS: 52.0

Experiment 4 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VPAND YMM0, YMM1, YMM1;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPCMPEQD YMM13, YMM13, YMM13;VPOR YMM0, YMM0, YMM13
Chain latency: 12

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c5 f5 db c1          	vpand  ymm0,ymm1,ymm1
   a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
   e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  12:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  16:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  22:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  26:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  32:	c4 41 15 76 ed       	vpcmpeqd ymm13,ymm13,ymm13
  37:	c4 c1 7d eb c5       	vpor   ymm0,ymm0,ymm13

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM1, [R14+YMM14], YMM0;
VGATHERDPS YMM0, [R14+YMM14], YMM1

Show nanoBench command
Results:
- MPERF: 25.28
- APERF: 31.0
- UOPS: 52.0

Latency operand 3 → 3: 0

Experiment 1

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VPCMPEQD YMM13, YMM13, YMM13;VPOR YMM0, YMM0, YMM13
Chain latency: 1

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c4 41 15 76 ed       	vpcmpeqd ymm13,ymm13,ymm13
   b:	c4 c1 7d eb c5       	vpor   ymm0,ymm0,ymm13

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 7.77
- APERF: 9.42
- UOPS: 41.0

Experiment 2

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VPCMPEQD YMM13, YMM13, YMM13;VPOR YMM0, YMM0, YMM13
Chain latency: 12

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
   a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
   e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  12:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  16:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  22:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  26:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  32:	c4 41 15 76 ed       	vpcmpeqd ymm13,ymm13,ymm13
  37:	c4 c1 7d eb c5       	vpor   ymm0,ymm0,ymm13

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 9.31
- APERF: 11.54
- UOPS: 52.0

Experiment 3

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPCMPEQD YMM13, YMM13, YMM13;VPOR YMM0, YMM0, YMM13
Chain latency: 12

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
   a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
   e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  12:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  16:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  22:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  26:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  32:	c4 41 15 76 ed       	vpcmpeqd ymm13,ymm13,ymm13
  37:	c4 c1 7d eb c5       	vpor   ymm0,ymm0,ymm13

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14

Show nanoBench command
Results:
- MPERF: 9.31
- APERF: 11.62
- UOPS: 52.0

Experiment 4 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VPCMPEQD YMM13, YMM13, YMM13;VPOR YMM0, YMM0, YMM13
Chain latency: 1

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c4 41 15 76 ed       	vpcmpeqd ymm13,ymm13,ymm13
   b:	c4 c1 7d eb c5       	vpor   ymm0,ymm0,ymm13

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM1, [R14+YMM14], YMM0;
VGATHERDPS YMM0, [R14+YMM14], YMM1

Show nanoBench command
Results:
- MPERF: 7.65
- APERF: 9.32
- UOPS: 41.0

Experiment 5 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VANDPD YMM0, YMM0, YMM0;VPCMPEQD YMM13, YMM13, YMM13;VPOR YMM0, YMM0, YMM13
Chain latency: 12

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
   a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
   e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  12:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  16:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  1e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  22:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  26:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2a:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  2e:	c5 fd 54 c0          	vandpd ymm0,ymm0,ymm0
  32:	c4 41 15 76 ed       	vpcmpeqd ymm13,ymm13,ymm13
  37:	c4 c1 7d eb c5       	vpor   ymm0,ymm0,ymm13

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM1, [R14+YMM14], YMM0;
VGATHERDPS YMM0, [R14+YMM14], YMM1

Show nanoBench command
Results:
- MPERF: 9.31
- APERF: 11.71
- UOPS: 52.0

Experiment 6 (source registers initialized by an instruction of the same kind)

Instruction: VGATHERDPS YMM1, [R14+YMM14], YMM0
Chain instructions: VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPAND YMM0, YMM0, YMM0;VPCMPEQD YMM13, YMM13, YMM13;VPOR YMM0, YMM0, YMM13
Chain latency: 12

Code:

   0:	c4 82 7d 92 0c 36    	vgatherdps ymm1,DWORD PTR [r14+ymm14*1],ymm0
   6:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
   a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
   e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  12:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  16:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  1e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  22:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  26:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2a:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  2e:	c5 fd db c0          	vpand  ymm0,ymm0,ymm0
  32:	c4 41 15 76 ed       	vpcmpeqd ymm13,ymm13,ymm13
  37:	c4 c1 7d eb c5       	vpor   ymm0,ymm0,ymm13

Init:

MOV R15, 10000;
L: VADDPS YMM0, YMM1, YMM1;
VADDPS YMM0, YMM1, YMM1;
DEC R15;
JNZ L;
VZEROALL;
MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOV [R14+16], RAX;
MOV [R14+24], RAX;
VMOVUPD YMM1, [R14];
VXORPS YMM0, YMM0, YMM0;
VXORPS YMM14, YMM14, YMM14;
VGATHERDPS YMM1, [R14+YMM14], YMM0;
VGATHERDPS YMM0, [R14+YMM14], YMM1

Show nanoBench command
Results:
- MPERF: 9.44
- APERF: 11.62
- UOPS: 52.0

VGATHERDPS (YMM, VSIB_YMM, YMM) - Latency

Operands

Latency operand 1 → 1: 8

Latency operand 1 → 3: 0

Latency operand 2 → 1 (address, base register): ≤20

Latency operand 2 → 1 (address, index register): ≤28

Latency operand 2 → 1 (memory): ≤16

Latency operand 2 → 3 (address, base register): ≤6

Latency operand 2 → 3 (address, index register): ≤1

Latency operand 2 → 3 (memory): ≤0

Latency operand 3 → 1: 19

Latency operand 3 → 3: 0

Latency operand 1 → 1: 8

Experiment 1

Experiment 2

Experiment 3

Experiment 4 (source registers initialized by an instruction of the same kind)

Experiment 5 (source registers initialized by an instruction of the same kind)

Experiment 6 (source registers initialized by an instruction of the same kind)

Latency operand 1 → 3: 0

Experiment 1

Experiment 2

Experiment 3 (source registers initialized by an instruction of the same kind)

Experiment 4 (source registers initialized by an instruction of the same kind)

Latency operand 2 → 1 (address, base register): ≤20

Experiment 1

Experiment 2 (source registers initialized by an instruction of the same kind)

Latency operand 2 → 1 (address, index register): ≤28

Experiment 1

Experiment 2 (source registers initialized by an instruction of the same kind)

Latency operand 2 → 1 (memory): ≤16

Experiment 1

Experiment 2

Experiment 3 (source registers initialized by an instruction of the same kind)

Experiment 4 (source registers initialized by an instruction of the same kind)

Latency operand 2 → 3 (address, base register): ≤6

Experiment 1

Experiment 2 (source registers initialized by an instruction of the same kind)

Latency operand 2 → 3 (address, index register): ≤1

Experiment 1

Experiment 2 (source registers initialized by an instruction of the same kind)

Latency operand 2 → 3 (memory): ≤0

Experiment 1

Experiment 2

Experiment 3 (source registers initialized by an instruction of the same kind)

Experiment 4 (source registers initialized by an instruction of the same kind)

Latency operand 3 → 1: 19

Experiment 1

Experiment 2

Experiment 3 (source registers initialized by an instruction of the same kind)

Experiment 4 (source registers initialized by an instruction of the same kind)

Latency operand 3 → 3: 0

Experiment 1

Experiment 2

Experiment 3

Experiment 4 (source registers initialized by an instruction of the same kind)

Experiment 5 (source registers initialized by an instruction of the same kind)

Experiment 6 (source registers initialized by an instruction of the same kind)