OV - attention-avx512 - Loop 50

	Loop Id: 50	Module: attention-avx512	Source: attention.cpp:27-33	Coverage: 1.12%

0x5840 VCVTSD2SS	%XMM1,%XMM1,%XMM1

0x5844 MOV	0xa0(%RSP),%RAX

0x584c MOV	0xb0(%RSP),%R13

0x5854 LEA	(%R9,%RDI,1),%R14D

0x5858 INC	%RDI

0x585b VMULSS	%XMM0,%XMM1,%XMM1

0x585f VMOVSS	%XMM1,(%RAX,%R14,4)

0x5865 CMP	%RDI,0x50(%RSP)

0x586a JE	57e0

0x5870 VXORPS	%XMM1,%XMM1,%XMM1

0x5874 CMP	$0x4,%R15D

0x5878 JB	5890

0x587a MOV	%EDI,%R14D

0x587d ADD	0x20(%RSP),%R14D

0x5882 SETB	%R14B

0x5886 OR	0xc0(%RSP),%R14B

0x588e JE	58e0

0x5890 MOV	0x90(%RSP),%R11

0x5898 MOV	0x60(%RSP),%RCX

0x589d MOV	0x38(%RSP),%RAX

0x58a2 XOR	%R14D,%R14D

0x58a5 MOV	%RAX,%R13

0x58a8 IMUL	%R14,%R13

0x58ac ADD	%RDI,%R13

0x58af NOP

  (47) 0x58b0 LEA	(%R10,%R14,1),%EDX

  (47) 0x58b4 INC	%R14

  (47) 0x58b7 VMOVSS	(%R11,%RDX,4),%XMM2

  (47) 0x58bd MOV	%R13D,%EDX

  (47) 0x58c0 VMOVSS	(%RCX,%RDX,4),%XMM3

  (47) 0x58c5 ADD	%RAX,%R13

  (47) 0x58c8 VCVTSS2SD	%XMM2,%XMM2,%XMM2

  (47) 0x58cc VCVTSS2SD	%XMM3,%XMM3,%XMM3

  (47) 0x58d0 VFMADD231SD	%XMM3,%XMM2,%XMM1

  (47) 0x58d5 CMP	%R14,%R15

  (47) 0x58d8 JNE	58b0

0x58da JMP	5840

0x58e0 CMP	$0x20,%R15D

0x58e4 JAE	58fb

0x58e6 MOV	0x90(%RSP),%R11

0x58ee MOV	0x60(%RSP),%RCX

0x58f3 XOR	%R13D,%R13D

0x58f6 JMP	59e3

0x58fb MOV	0x90(%RSP),%RDX

0x5903 MOV	0x60(%RSP),%RCX

0x5908 MOV	0x40(%RSP),%RAX

0x590d VXORPS	%XMM1,%XMM1,%XMM1

0x5911 VXORPS	%XMM2,%XMM2,%XMM2

0x5915 VPXOR	%XMM3,%XMM3,%XMM3

0x5919 VPXOR	%XMM4,%XMM4,%XMM4

0x591d XOR	%R14D,%R14D

  (51) 0x5920 LEA	(%R10,%R14,1),%R13D

  (51) 0x5924 VCVTPS2PD	(%RDX,%R13,4),%ZMM5

  (51) 0x592b VCVTPS2PD	0x20(%RDX,%R13,4),%ZMM6

  (51) 0x5933 VCVTPS2PD	0x40(%RDX,%R13,4),%ZMM7

  (51) 0x593b VCVTPS2PD	0x60(%RDX,%R13,4),%ZMM8

  (51) 0x5943 LEA	(%RDI,%R14,1),%R13D

  (51) 0x5947 ADD	$0x20,%R14

  (51) 0x594b VCVTPS2PD	(%RCX,%R13,4),%ZMM9

  (51) 0x5952 VCVTPS2PD	0x20(%RCX,%R13,4),%ZMM10

  (51) 0x595a VFMADD231PD	%ZMM9,%ZMM5,%ZMM1

  (51) 0x5960 VCVTPS2PD	0x40(%RCX,%R13,4),%ZMM9

  (51) 0x5968 VFMADD231PD	%ZMM10,%ZMM6,%ZMM2

  (51) 0x596e VCVTPS2PD	0x60(%RCX,%R13,4),%ZMM6

  (51) 0x5976 VFMADD231PD	%ZMM9,%ZMM7,%ZMM3

  (51) 0x597c VFMADD231PD	%ZMM6,%ZMM8,%ZMM4

  (51) 0x5982 CMP	%R14,%RSI

  (51) 0x5985 JNE	5920

0x5987 VADDPD	%ZMM1,%ZMM2,%ZMM1

0x598d VADDPD	%ZMM3,%ZMM4,%ZMM3

0x5993 VADDPD	%ZMM1,%ZMM3,%ZMM1

0x5999 VEXTRACTF64X4	$0x1,%ZMM1,%YMM2

0x59a0 VADDPD	%ZMM2,%ZMM1,%ZMM1

0x59a6 VEXTRACTF128	$0x1,%YMM1,%XMM2

0x59ac VADDPD	%XMM2,%XMM1,%XMM1

0x59b0 VSHUFPD	$0x1,%XMM1,%XMM1,%XMM2

0x59b5 VADDSD	%XMM2,%XMM1,%XMM1

0x59b9 TEST	%RAX,%RAX

0x59bc JE	5840

0x59c2 MOV	%RSI,%R13

0x59c5 MOV	%RSI,%R14

0x59c8 MOV	0x90(%RSP),%R11

0x59d0 MOV	0x60(%RSP),%RCX

0x59d5 CMP	$0x4,%EAX

0x59d8 MOV	0x38(%RSP),%RAX

0x59dd JB	58a5

0x59e3 VMOVQ	%XMM1,%XMM1

0x59e7 NOPW	(%RAX,%RAX,1)

  (52) 0x59f0 LEA	(%R10,%R13,1),%R14D

  (52) 0x59f4 VCVTPS2PD	(%R11,%R14,4),%YMM2

  (52) 0x59fa LEA	(%RDI,%R13,1),%R14D

  (52) 0x59fe ADD	$0x4,%R13

  (52) 0x5a02 VCVTPS2PD	(%RCX,%R14,4),%YMM3

  (52) 0x5a08 VFMADD231PD	%YMM3,%YMM2,%YMM1

  (52) 0x5a0d CMP	%R13,%R8

  (52) 0x5a10 JNE	59f0

0x5a12 VEXTRACTF128	$0x1,%YMM1,%XMM2

0x5a18 CMPQ	$0,0x30(%RSP)

0x5a1e MOV	0x38(%RSP),%RAX

0x5a23 MOV	%R8,%R14

0x5a26 VADDPD	%XMM2,%XMM1,%XMM1

0x5a2a VSHUFPD	$0x1,%XMM1,%XMM1,%XMM2

0x5a2f VADDSD	%XMM2,%XMM1,%XMM1

0x5a33 JNE	58a5

0x5a39 JMP	5840

/home/eoseret/llm-attention/attention.cpp: 27 - 33

--------------------------------------------------------------------------------

27:       for (unsigned int j = 0; j < N; ++j) {  //vectorized

28:          double sum = 0.0;

29: #pragma clang loop vectorize(enable)

30:          for (unsigned int k = 0; k < K; ++k) { //vectorized

31:             sum += (double)A[i * K + k] * (double)B[k * N + j];

32:          }

33:          C[i * N + j] = alpha * static_cast<float>(sum);

Coverage (%)	Name	Source Location	Module

min	med	avg	max

Percentile Index	10	20	30	40	50	60	70	80	90	100
Value

min	med	avg	max

Percentile Index	10	20	30	40	50	60	70	80	90	100
Value

Path /

Metric	Value
CQA speedup if no scalar integer	1.86
CQA speedup if FP arith vectorized	1.06
CQA speedup if fully vectorized	1.52
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.76
Bottlenecks	micro-operation queue,
Function	main
Source	attention.cpp:27-27,attention.cpp:30-30,attention.cpp:33-33
Source loop unroll info	NA
Source loop unroll confidence level	NA
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	8.38
CQA cycles if no scalar integer	4.50
CQA cycles if FP arith vectorized	7.87
CQA cycles if fully vectorized	5.50
Front-end cycles	8.38
P0 cycles	3.83
P1 cycles	3.83
P2 cycles	3.83
P3 cycles	3.83
P4 cycles	3.83
P5 cycles	3.83
P6 cycles	4.75
P7 cycles	4.75
P8 cycles	4.75
P9 cycles	4.75
P10 cycles	3.50
P11 cycles	3.50
P12 cycles	4.50
P13 cycles	4.50
P14 cycles	0.50
P15 cycles	0.50
DIV/SQRT cycles	0.00
Inter-iter dependencies cycles	NA
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	69.00
Nb uops	67.00
Nb loads	18.00
Nb stores	1.00
Nb stack references	10.00
FLOP/cycle	4.66
Nb FLOP add-sub	38.00
Nb FLOP mul	1.00
Nb FLOP fma	0.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	16.36
Bytes prefetched	0.00
Bytes loaded	133.00
Bytes stored	4.00
Stride 0	NA
Stride 1	NA
Stride n	NA
Stride unknown	NA
Stride indirect	NA
Vectorization ratio all	39.02
Vectorization ratio load	0.00
Vectorization ratio store	0.00
Vectorization ratio mul	0.00
Vectorization ratio add_sub	66.67
Vectorization ratio fma	NA
Vectorization ratio div_sqrt	NA
Vectorization ratio other	50.00
Vector-efficiency ratio all	24.39
Vector-efficiency ratio load	12.50
Vector-efficiency ratio store	6.25
Vector-efficiency ratio mul	9.38
Vector-efficiency ratio add_sub	54.17
Vector-efficiency ratio fma	NA
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	18.75

Metric	Value
CQA speedup if no scalar integer	1.86
CQA speedup if FP arith vectorized	1.06
CQA speedup if fully vectorized	1.52
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.76
Bottlenecks	micro-operation queue,
Function	main
Source	attention.cpp:27-27,attention.cpp:30-30,attention.cpp:33-33
Source loop unroll info	NA
Source loop unroll confidence level	NA
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	8.38
CQA cycles if no scalar integer	4.50
CQA cycles if FP arith vectorized	7.87
CQA cycles if fully vectorized	5.50
Front-end cycles	8.38
P0 cycles	3.83
P1 cycles	3.83
P2 cycles	3.83
P3 cycles	3.83
P4 cycles	3.83
P5 cycles	3.83
P6 cycles	4.75
P7 cycles	4.75
P8 cycles	4.75
P9 cycles	4.75
P10 cycles	3.50
P11 cycles	3.50
P12 cycles	4.50
P13 cycles	4.50
P14 cycles	0.50
P15 cycles	0.50
DIV/SQRT cycles	0.00
Inter-iter dependencies cycles	NA
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	69.00
Nb uops	67.00
Nb loads	18.00
Nb stores	1.00
Nb stack references	10.00
FLOP/cycle	4.66
Nb FLOP add-sub	38.00
Nb FLOP mul	1.00
Nb FLOP fma	0.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	16.36
Bytes prefetched	0.00
Bytes loaded	133.00
Bytes stored	4.00
Stride 0	NA
Stride 1	NA
Stride n	NA
Stride unknown	NA
Stride indirect	NA
Vectorization ratio all	39.02
Vectorization ratio load	0.00
Vectorization ratio store	0.00
Vectorization ratio mul	0.00
Vectorization ratio add_sub	66.67
Vectorization ratio fma	NA
Vectorization ratio div_sqrt	NA
Vectorization ratio other	50.00
Vector-efficiency ratio all	24.39
Vector-efficiency ratio load	12.50
Vector-efficiency ratio store	6.25
Vector-efficiency ratio mul	9.38
Vector-efficiency ratio add_sub	54.17
Vector-efficiency ratio fma	NA
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	18.75

Path /

Average path: Display a virtual path defined by average values of all real paths

Function	main
Source file and lines	attention.cpp:27-33
Module	attention-avx512

The loop is defined in /home/eoseret/llm-attention/attention.cpp:27-33.

Analyzed code is defined in /home/eoseret/llm-attention/attention.cpp:27,30,33.

Warnings:

Non-innermost loop: analyzing only self part (ignoring child loops).
Ignoring paths for analysis
Too many paths. Rerun with max-paths=8
RecMII not computed since number of paths is unknown or > max_paths
Streams not analyzed since number of paths is unknown or > max_paths

Try to simplify control and/or increase the maximum number of paths per function/loop through the 'max-paths-nb' option.

This loop has 8 execution paths.

The presence of multiple execution paths is typically the main/first bottleneck.
Try to simplify control inside loop: ideally, try to remove all conditional expressions, for example by (if applicable):

hoisting them (moving them outside the loop)
turning them into conditional moves, MIN or MAX

Ex: if (x<0) x=0 => x = (x<0 ? 0 : x) (or MAX(0,x) after defining the corresponding macro)

gain
potential
hint
expert

Code clean check

Detected a slowdown caused by scalar integer instructions (typically used for address computation). By removing them, you can lower the cost of an iteration from 8.38 to 4.50 cycles (1.86x speedup).

Workaround

Try to reorganize arrays of structures to structures of arrays
Consider to permute loops (see vectorization gain report)

Vectorization

Your loop is poorly vectorized. Only 24% of vector register length is used (average across all SSE/AVX instructions). By fully vectorizing your loop, you can lower the cost of an iteration from 8.38 to 5.50 cycles (1.52x speedup).

Details

39% of SSE/AVX instructions are used in vector version (process two or more data elements in vector registers):

0% of SSE/AVX loads are used in vector version.
0% of SSE/AVX stores are used in vector version.
66% of SSE/AVX addition or subtraction instructions are used in vector version.
0% of SSE/AVX multiply instructions are used in vector version.
50% of SSE/AVX instructions that are not load, store, addition, subtraction nor multiply instructions are used in vector version.

Since your execution units are vector units, only a fully vectorized loop can use their full power.

Workaround

Try another compiler or update/tune your current one
Remove inter-iterations dependences from your loop and make it unit-stride:
- If your arrays have 2 or more dimensions, check whether elements are accessed contiguously and, otherwise, try to permute loops accordingly: C storage order is row-major: for(i) for(j) a[j][i] = b[j][i]; (slow, non stride 1) => for(i) for(j) a[i][j] = b[i][j]; (fast, stride 1)
- If your loop streams arrays of structures (AoS), try to use structures of arrays instead (SoA): for(i) a[i].x = b[i].x; (slow, non stride 1) => for(i) a.x[i] = b.x[i]; (fast, stride 1)

Execution units bottlenecks

Found no such bottlenecks but see expert reports for more complex bottlenecks.

FMA

Presence of both ADD/SUB and MUL operations.

Workaround

Pass to your compiler a micro-architecture specialization option:
- use march=native
Try to change order in which elements are evaluated (using parentheses) in arithmetic expressions containing both ADD/SUB and MUL operations to enable your compiler to generate FMA instructions wherever possible. For instance a + b*c is a valid FMA (MUL then ADD). However (a+b)* c cannot be translated into an FMA (ADD then MUL).

Vector unaligned load/store instructions

Detected 3 suboptimal vector unaligned load/store instructions.

Details

VEXTRACTF128: 2 occurrences►
- /home/eoseret/llm-attention/attention.cpp:30
- /home/eoseret/llm-attention/attention.cpp:30
VEXTRACTF64X4: 1 occurrences►
- /home/eoseret/llm-attention/attention.cpp:30

Workaround

Pass to your compiler a micro-architecture specialization option:
- use march=native
Use vector aligned instructions:
1. align your arrays on 64 bytes boundaries: replace { void *p = malloc (size); } with { void *p; posix_memalign (&p, 64, size); }.
2. inform your compiler that your arrays are vector aligned: if array 'foo' is 64 bytes-aligned, define a pointer 'p_foo' as __builtin_assume_aligned (foo, 64) and use it instead of 'foo' in the loop.

Conversion instructions

Detected expensive conversion instructions.

Details

VCVTSD2SS (FP64 to FP32, scalar): 1 occurrences►
- /home/eoseret/llm-attention/attention.cpp:33

Workaround

Avoid mixing data with different types. In particular, check if the type of constants is the same as array elements. Use double instead of single precision only when/where needed by numerical stability and avoid mixing precision. In C/C++, FP constants are double precision by default and must be suffixed by 'f' to make them single precision.

Type of elements and instruction set

1 SSE or AVX instructions are processing arithmetic or math operations on single precision FP elements in scalar mode (one at a time). 2 SSE or AVX instructions are processing arithmetic or math operations on double precision FP elements in scalar mode (one at a time). 2 SSE or AVX instructions are processing arithmetic or math operations on double precision FP elements in vector mode (two at a time). 4 AVX-512 instructions are processing arithmetic or math operations on double precision FP elements in vector mode (eight at a time).

Matching between your loop (in the source code) and the binary loop

The binary loop is composed of 39 FP arithmetical operations:

38: addition or subtraction
1: multiply

The binary loop is loading 133 bytes (16 double precision FP elements). The binary loop is storing 4 bytes (0 double precision FP elements).

Arithmetic intensity

Arithmetic intensity is 0.28 FP operations per loaded or stored byte.

General properties

nb instructions	69
nb uops	67
loop length	330
used x86 registers	12
used mmx registers	0
used xmm registers	5
used ymm registers	2
used zmm registers	4
nb stack references	10
ADD-SUB / MUL ratio	8.00

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	8.38 cycles
front end	8.38 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15
uops	3.83	3.83	3.83	3.83	3.83	3.83	4.75	4.75	4.75	4.75	3.50	3.50	4.50	4.50	0.50	0.50
cycles	3.83	3.83	3.83	3.83	3.83	3.83	4.75	4.75	4.75	4.75	3.50	3.50	4.50	4.50	0.50	0.50

Execution ports to units layout:

P0: ALU
P1: ALU
P2: ALU
P3: ALU, BRU
P4: ALU, BRU
P5: ALU, BRU
P6 (512 bits): store address, load
P7 (512 bits): store address, load
P8 (512 bits): store address, load
P9 (512 bits): store address, load
P10 (512 bits): VPU, DIV/SQRT
P11 (512 bits): VPU, DIV/SQRT
P12 (512 bits): VPU
P13 (512 bits): VPU
P14 (512 bits): FP store data
P15 (512 bits): FP store data

Cycles executing div or sqrt instructions

Cycles summary

Front-end	8.38
Dispatch	4.75
Overall L1	8.38

Vectorization ratios

INT

all	9%
load	0%
store	NA (no store vectorizable/vectorized instructions)
mul	0%
add-sub	0%
fma	NA (no fma vectorizable/vectorized instructions)
other	18%

all	73%
load	NA (no load vectorizable/vectorized instructions)
store	0%
mul	0%
add-sub	75%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	88%

INT+FP

all	39%
load	0%
store	0%
mul	0%
add-sub	66%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	50%

Vector efficiency ratios

INT

all	12%
load	12%
store	NA (no store vectorizable/vectorized instructions)
mul	12%
add-sub	12%
fma	NA (no fma vectorizable/vectorized instructions)
other	12%

all	38%
load	NA (no load vectorizable/vectorized instructions)
store	6%
mul	6%
add-sub	59%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	26%

INT+FP

all	24%
load	12%
store	6%
mul	9%
add-sub	54%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	18%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 8.38 cycles. At this rate:

6% of peak load performance is reached (15.88 out of 256.00 bytes loaded per cycle (GB/s @ 1GHz))
0% of peak store performance is reached (0.48 out of 128.00 bytes stored per cycle (GB/s @ 1GHz))

Front-end bottlenecks

Performance is limited by instruction throughput (loading/decoding program instructions to execution core) (front-end is a bottleneck). By removing all these bottlenecks, you can lower the cost of an iteration from 8.38 to 4.75 cycles (1.76x speedup).

ASM code

In the binary file, the address of the loop is: 5870

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15	Latency	Recip. throughput	Vectorization
VCVTSD2SS %XMM1,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	scal (12.5%)
MOV 0xa0(%RSP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	N/A
MOV 0xb0(%RSP),%R13	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
LEA (%R9,%RDI,1),%R14D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
INC %RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
VMULSS %XMM0,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (6.3%)
VMOVSS %XMM1,(%RAX,%R14,4)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	1	0.50	scal (6.3%)
CMP %RDI,0x50(%RSP)	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 57e0 <main+0x1f00>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.33-1	N/A
VXORPS %XMM1,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
CMP $0x4,%R15D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	scal (6.3%)
JB 5890 <main+0x1fb0>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.33-1	N/A
MOV %EDI,%R14D	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
ADD 0x20(%RSP),%R14D	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	0	0	0	0	1	0.25	N/A
SETB %R14B	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
OR 0xc0(%RSP),%R14B	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	0	0	0	0	1	0.25	N/A
JE 58e0 <main+0x2000>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.33-1	N/A
MOV 0x90(%RSP),%R11	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
MOV 0x60(%RSP),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
MOV 0x38(%RSP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	N/A
XOR %R14D,%R14D	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV %RAX,%R13	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	scal (12.5%)
IMUL %R14,%R13	1	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0	0	3	0.33	scal (12.5%)
ADD %RDI,%R13	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	scal (12.5%)
NOP	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
JMP 5840 <main+0x1f60>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.50	N/A
CMP $0x20,%R15D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	scal (6.3%)
JAE 58fb <main+0x201b>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.33-1	N/A
MOV 0x90(%RSP),%R11	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
MOV 0x60(%RSP),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
XOR %R13D,%R13D	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	scal (6.3%)
JMP 59e3 <main+0x2103>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.50	N/A
MOV 0x90(%RSP),%RDX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
MOV 0x60(%RSP),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
MOV 0x40(%RSP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	N/A
VXORPS %XMM1,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
VXORPS %XMM2,%XMM2,%XMM2	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
VPXOR %XMM3,%XMM3,%XMM3	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
VPXOR %XMM4,%XMM4,%XMM4	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
XOR %R14D,%R14D	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
VADDPD %ZMM1,%ZMM2,%ZMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	vect (100.0%)
VADDPD %ZMM3,%ZMM4,%ZMM3	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	vect (100.0%)
VADDPD %ZMM1,%ZMM3,%ZMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	vect (100.0%)
VEXTRACTF64X4 $0x1,%ZMM1,%YMM2	1	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	4	0.50	vect (50.0%)
VADDPD %ZMM2,%ZMM1,%ZMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	vect (100.0%)
VEXTRACTF128 $0x1,%YMM1,%XMM2	1	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	3	0.50	vect (25.0%)
VADDPD %XMM2,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	vect (25.0%)
VSHUFPD $0x1,%XMM1,%XMM1,%XMM2	1	0	0	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	1	0.25	vect (25.0%)
VADDSD %XMM2,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
TEST %RAX,%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	scal (12.5%)
JE 5840 <main+0x1f60>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.33-1	N/A
MOV %RSI,%R13	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	scal (12.5%)
MOV %RSI,%R14	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV 0x90(%RSP),%R11	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
MOV 0x60(%RSP),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
CMP $0x4,%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	scal (6.3%)
MOV 0x38(%RSP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	N/A
JB 58a5 <main+0x1fc5>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.33-1	N/A
VMOVQ %XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	1	0.25	scal (12.5%)
NOPW (%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.13	N/A
VEXTRACTF128 $0x1,%YMM1,%XMM2	1	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	3	0.50	vect (25.0%)
CMPQ $0,0x30(%RSP)	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	0	0	0	0	1	0.25	N/A
MOV 0x38(%RSP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	N/A
MOV %R8,%R14	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
VADDPD %XMM2,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	vect (25.0%)
VSHUFPD $0x1,%XMM1,%XMM1,%XMM2	1	0	0	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	1	0.25	vect (25.0%)
VADDSD %XMM2,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
JNE 58a5 <main+0x1fc5>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.33-1	N/A
JMP 5840 <main+0x1f60>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.50	N/A

Function	main
Source file and lines	attention.cpp:27-33
Module	attention-avx512

The loop is defined in /home/eoseret/llm-attention/attention.cpp:27-33.

Analyzed code is defined in /home/eoseret/llm-attention/attention.cpp:27,30,33.

Warnings:

Non-innermost loop: analyzing only self part (ignoring child loops).
Ignoring paths for analysis
Too many paths. Rerun with max-paths=8
RecMII not computed since number of paths is unknown or > max_paths
Streams not analyzed since number of paths is unknown or > max_paths

hoisting them (moving them outside the loop)
turning them into conditional moves, MIN or MAX

Ex: if (x<0) x=0 => x = (x<0 ? 0 : x) (or MAX(0,x) after defining the corresponding macro)

gain
potential
hint
expert

Code clean check

Detected a slowdown caused by scalar integer instructions (typically used for address computation). By removing them, you can lower the cost of an iteration from 8.38 to 4.50 cycles (1.86x speedup).

Workaround

Try to reorganize arrays of structures to structures of arrays
Consider to permute loops (see vectorization gain report)

Vectorization

Details

39% of SSE/AVX instructions are used in vector version (process two or more data elements in vector registers):

0% of SSE/AVX loads are used in vector version.
0% of SSE/AVX stores are used in vector version.
66% of SSE/AVX addition or subtraction instructions are used in vector version.
0% of SSE/AVX multiply instructions are used in vector version.
50% of SSE/AVX instructions that are not load, store, addition, subtraction nor multiply instructions are used in vector version.

Since your execution units are vector units, only a fully vectorized loop can use their full power.

Workaround

Try another compiler or update/tune your current one
Remove inter-iterations dependences from your loop and make it unit-stride:
- If your arrays have 2 or more dimensions, check whether elements are accessed contiguously and, otherwise, try to permute loops accordingly: C storage order is row-major: for(i) for(j) a[j][i] = b[j][i]; (slow, non stride 1) => for(i) for(j) a[i][j] = b[i][j]; (fast, stride 1)
- If your loop streams arrays of structures (AoS), try to use structures of arrays instead (SoA): for(i) a[i].x = b[i].x; (slow, non stride 1) => for(i) a.x[i] = b.x[i]; (fast, stride 1)

Execution units bottlenecks

Found no such bottlenecks but see expert reports for more complex bottlenecks.

FMA

Presence of both ADD/SUB and MUL operations.

Workaround

Pass to your compiler a micro-architecture specialization option:
- use march=native
Try to change order in which elements are evaluated (using parentheses) in arithmetic expressions containing both ADD/SUB and MUL operations to enable your compiler to generate FMA instructions wherever possible. For instance a + b*c is a valid FMA (MUL then ADD). However (a+b)* c cannot be translated into an FMA (ADD then MUL).

Vector unaligned load/store instructions

Detected 3 suboptimal vector unaligned load/store instructions.

Details

VEXTRACTF128: 2 occurrences►
- /home/eoseret/llm-attention/attention.cpp:30
- /home/eoseret/llm-attention/attention.cpp:30
VEXTRACTF64X4: 1 occurrences►
- /home/eoseret/llm-attention/attention.cpp:30

Workaround

Pass to your compiler a micro-architecture specialization option:
- use march=native
Use vector aligned instructions:
1. align your arrays on 64 bytes boundaries: replace { void *p = malloc (size); } with { void *p; posix_memalign (&p, 64, size); }.
2. inform your compiler that your arrays are vector aligned: if array 'foo' is 64 bytes-aligned, define a pointer 'p_foo' as __builtin_assume_aligned (foo, 64) and use it instead of 'foo' in the loop.

Conversion instructions

Detected expensive conversion instructions.

Details

VCVTSD2SS (FP64 to FP32, scalar): 1 occurrences►
- /home/eoseret/llm-attention/attention.cpp:33

Workaround

Type of elements and instruction set

Matching between your loop (in the source code) and the binary loop

The binary loop is composed of 39 FP arithmetical operations:

38: addition or subtraction
1: multiply

The binary loop is loading 133 bytes (16 double precision FP elements). The binary loop is storing 4 bytes (0 double precision FP elements).

Arithmetic intensity

Arithmetic intensity is 0.28 FP operations per loaded or stored byte.

General properties

nb instructions	69
nb uops	67
loop length	330
used x86 registers	12
used mmx registers	0
used xmm registers	5
used ymm registers	2
used zmm registers	4
nb stack references	10
ADD-SUB / MUL ratio	8.00

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	8.38 cycles
front end	8.38 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15
uops	3.83	3.83	3.83	3.83	3.83	3.83	4.75	4.75	4.75	4.75	3.50	3.50	4.50	4.50	0.50	0.50
cycles	3.83	3.83	3.83	3.83	3.83	3.83	4.75	4.75	4.75	4.75	3.50	3.50	4.50	4.50	0.50	0.50

Execution ports to units layout:

P0: ALU
P1: ALU
P2: ALU
P3: ALU, BRU
P4: ALU, BRU
P5: ALU, BRU
P6 (512 bits): store address, load
P7 (512 bits): store address, load
P8 (512 bits): store address, load
P9 (512 bits): store address, load
P10 (512 bits): VPU, DIV/SQRT
P11 (512 bits): VPU, DIV/SQRT
P12 (512 bits): VPU
P13 (512 bits): VPU
P14 (512 bits): FP store data
P15 (512 bits): FP store data

Cycles executing div or sqrt instructions

Cycles summary

Front-end	8.38
Dispatch	4.75
Overall L1	8.38

Vectorization ratios

INT

all	9%
load	0%
store	NA (no store vectorizable/vectorized instructions)
mul	0%
add-sub	0%
fma	NA (no fma vectorizable/vectorized instructions)
other	18%

all	73%
load	NA (no load vectorizable/vectorized instructions)
store	0%
mul	0%
add-sub	75%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	88%

INT+FP

all	39%
load	0%
store	0%
mul	0%
add-sub	66%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	50%

Vector efficiency ratios

INT

all	12%
load	12%
store	NA (no store vectorizable/vectorized instructions)
mul	12%
add-sub	12%
fma	NA (no fma vectorizable/vectorized instructions)
other	12%

all	38%
load	NA (no load vectorizable/vectorized instructions)
store	6%
mul	6%
add-sub	59%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	26%

INT+FP

all	24%
load	12%
store	6%
mul	9%
add-sub	54%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	18%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 8.38 cycles. At this rate:

6% of peak load performance is reached (15.88 out of 256.00 bytes loaded per cycle (GB/s @ 1GHz))
0% of peak store performance is reached (0.48 out of 128.00 bytes stored per cycle (GB/s @ 1GHz))

Front-end bottlenecks

ASM code

In the binary file, the address of the loop is: 5870

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15	Latency	Recip. throughput	Vectorization
VCVTSD2SS %XMM1,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	scal (12.5%)
MOV 0xa0(%RSP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	N/A
MOV 0xb0(%RSP),%R13	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
LEA (%R9,%RDI,1),%R14D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
INC %RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
VMULSS %XMM0,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (6.3%)
VMOVSS %XMM1,(%RAX,%R14,4)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	1	0.50	scal (6.3%)
CMP %RDI,0x50(%RSP)	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 57e0 <main+0x1f00>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.33-1	N/A
VXORPS %XMM1,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
CMP $0x4,%R15D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	scal (6.3%)
JB 5890 <main+0x1fb0>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.33-1	N/A
MOV %EDI,%R14D	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
ADD 0x20(%RSP),%R14D	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	0	0	0	0	1	0.25	N/A
SETB %R14B	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
OR 0xc0(%RSP),%R14B	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	0	0	0	0	1	0.25	N/A
JE 58e0 <main+0x2000>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.33-1	N/A
MOV 0x90(%RSP),%R11	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
MOV 0x60(%RSP),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
MOV 0x38(%RSP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	N/A
XOR %R14D,%R14D	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV %RAX,%R13	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	scal (12.5%)
IMUL %R14,%R13	1	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0	0	3	0.33	scal (12.5%)
ADD %RDI,%R13	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	scal (12.5%)
NOP	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
JMP 5840 <main+0x1f60>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.50	N/A
CMP $0x20,%R15D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	scal (6.3%)
JAE 58fb <main+0x201b>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.33-1	N/A
MOV 0x90(%RSP),%R11	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
MOV 0x60(%RSP),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
XOR %R13D,%R13D	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	scal (6.3%)
JMP 59e3 <main+0x2103>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.50	N/A
MOV 0x90(%RSP),%RDX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
MOV 0x60(%RSP),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
MOV 0x40(%RSP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	N/A
VXORPS %XMM1,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
VXORPS %XMM2,%XMM2,%XMM2	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
VPXOR %XMM3,%XMM3,%XMM3	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
VPXOR %XMM4,%XMM4,%XMM4	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
XOR %R14D,%R14D	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
VADDPD %ZMM1,%ZMM2,%ZMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	vect (100.0%)
VADDPD %ZMM3,%ZMM4,%ZMM3	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	vect (100.0%)
VADDPD %ZMM1,%ZMM3,%ZMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	vect (100.0%)
VEXTRACTF64X4 $0x1,%ZMM1,%YMM2	1	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	4	0.50	vect (50.0%)
VADDPD %ZMM2,%ZMM1,%ZMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	vect (100.0%)
VEXTRACTF128 $0x1,%YMM1,%XMM2	1	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	3	0.50	vect (25.0%)
VADDPD %XMM2,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	vect (25.0%)
VSHUFPD $0x1,%XMM1,%XMM1,%XMM2	1	0	0	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	1	0.25	vect (25.0%)
VADDSD %XMM2,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
TEST %RAX,%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	scal (12.5%)
JE 5840 <main+0x1f60>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.33-1	N/A
MOV %RSI,%R13	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	scal (12.5%)
MOV %RSI,%R14	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV 0x90(%RSP),%R11	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
MOV 0x60(%RSP),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	scal (12.5%)
CMP $0x4,%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	scal (6.3%)
MOV 0x38(%RSP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	N/A
JB 58a5 <main+0x1fc5>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.33-1	N/A
VMOVQ %XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	1	0.25	scal (12.5%)
NOPW (%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.13	N/A
VEXTRACTF128 $0x1,%YMM1,%XMM2	1	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	3	0.50	vect (25.0%)
CMPQ $0,0x30(%RSP)	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	0	0	0	0	1	0.25	N/A
MOV 0x38(%RSP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	3	0.25	N/A
MOV %R8,%R14	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
VADDPD %XMM2,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	vect (25.0%)
VSHUFPD $0x1,%XMM1,%XMM1,%XMM2	1	0	0	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	1	0.25	vect (25.0%)
VADDSD %XMM2,%XMM1,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
JNE 58a5 <main+0x1fc5>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.33-1	N/A
JMP 5840 <main+0x1f60>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	0	0.50	N/A

Report Configuration

Code clean check

Workaround

Vectorization

Details

Workaround

Execution units bottlenecks

FMA

Workaround

Vector unaligned load/store instructions

Details

Workaround

Conversion instructions

Details

Workaround

Type of elements and instruction set

Matching between your loop (in the source code) and the binary loop

Arithmetic intensity

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code

Code clean check

Workaround

Vectorization

Details

Workaround

Execution units bottlenecks

FMA

Workaround

Vector unaligned load/store instructions

Details

Workaround

Conversion instructions

Details

Workaround

Type of elements and instruction set

Matching between your loop (in the source code) and the binary loop

Arithmetic intensity

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code