OV - exec - Loop 6

	Loop Id: 6	Module: exec	Source: stream.c:346-348	Coverage: 0.00%

0x401400 MADD	X16, X2, X0, X3

0x401404 ADD	X12, X2, X16

0x401408 CMP	X16, X12

0x40140c B.GE	401598

0x401410 ADRP	X4, 401410

0x401414 SUB	X5, X12, X16

0x401418 LDR	X6, [X4, #1456]

0x40141c ANDS	X7, X5, #4224

0x401420 LDR	X15, [X6, #16]

0x401424 LDP	X13, X14, [X6]

0x401428 B.EQ	4014f0

0x40142c CMP	X7, #1

0x401430 B.EQ	4014d4

0x401434 CMP	X7, #2

0x401438 B.EQ	4014c0

0x40143c CMP	X7, #3

0x401440 B.EQ	4014ac

0x401444 CMP	X7, #4

0x401448 B.EQ	401498

0x40144c CMP	X7, #5

0x401450 B.EQ	401484

0x401454 CMP	X7, #6

0x401458 B.EQ	401470

0x40145c LDR	D31, [X15, X16,LSL #3]

0x401460 LDR	D30, [X13, X16,LSL #3]

0x401464 FADD	D0, D31, D30

0x401468 STR	D0, [X14, X16,LSL #3]

0x40146c ADD	X16, X16, #1

0x401470 LDR	D29, [X15, X16,LSL #3]

0x401474 LDR	D28, [X13, X16,LSL #3]

0x401478 FADD	D1, D29, D28

0x40147c STR	D1, [X14, X16,LSL #3]

0x401480 ADD	X16, X16, #1

0x401484 LDR	D27, [X15, X16,LSL #3]

0x401488 LDR	D26, [X13, X16,LSL #3]

0x40148c FADD	D2, D27, D26

0x401490 STR	D2, [X14, X16,LSL #3]

0x401494 ADD	X16, X16, #1

0x401498 LDR	D25, [X15, X16,LSL #3]

0x40149c LDR	D24, [X13, X16,LSL #3]

0x4014a0 FADD	D3, D25, D24

0x4014a4 STR	D3, [X14, X16,LSL #3]

0x4014a8 ADD	X16, X16, #1

0x4014ac LDR	D23, [X15, X16,LSL #3]

0x4014b0 LDR	D22, [X13, X16,LSL #3]

0x4014b4 FADD	D4, D23, D22

0x4014b8 STR	D4, [X14, X16,LSL #3]

0x4014bc ADD	X16, X16, #1

0x4014c0 LDR	D21, [X15, X16,LSL #3]

0x4014c4 LDR	D20, [X13, X16,LSL #3]

0x4014c8 FADD	D5, D21, D20

0x4014cc STR	D5, [X14, X16,LSL #3]

0x4014d0 ADD	X16, X16, #1

0x4014d4 LDR	D19, [X15, X16,LSL #3]

0x4014d8 LDR	D18, [X13, X16,LSL #3]

0x4014dc FADD	D6, D19, D18

0x4014e0 STR	D6, [X14, X16,LSL #3]

0x4014e4 ADD	X16, X16, #1

0x4014e8 CMP	X12, X16

0x4014ec B.EQ	401598

  (7) 0x4014f0 LDR	D7, [X15, X16,LSL #3]

  (7) 0x4014f4 ADD	X17, X16, #1

  (7) 0x4014f8 ADD	X11, X16, #2

  (7) 0x4014fc ADD	X10, X16, #3

  (7) 0x401500 ADD	X9, X16, #4

  (7) 0x401504 LDR	D16, [X13, X16,LSL #3]

  (7) 0x401508 ADD	X8, X16, #5

  (7) 0x40150c ADD	X18, X16, #6

  (7) 0x401510 ADD	X30, X16, #7

  (7) 0x401514 FADD	D17, D7, D16

  (7) 0x401518 STR	D17, [X14, X16,LSL #3]

  (7) 0x40151c ADD	X16, X16, #8

  (7) 0x401520 LDR	D31, [X15, X17,LSL #3]

  (7) 0x401524 LDR	D30, [X13, X17,LSL #3]

  (7) 0x401528 FADD	D0, D31, D30

  (7) 0x40152c STR	D0, [X14, X17,LSL #3]

  (7) 0x401530 LDR	D29, [X15, X11,LSL #3]

  (7) 0x401534 LDR	D28, [X13, X11,LSL #3]

  (7) 0x401538 FADD	D1, D29, D28

  (7) 0x40153c STR	D1, [X14, X11,LSL #3]

  (7) 0x401540 LDR	D27, [X15, X10,LSL #3]

  (7) 0x401544 LDR	D26, [X13, X10,LSL #3]

  (7) 0x401548 FADD	D2, D27, D26

  (7) 0x40154c STR	D2, [X14, X10,LSL #3]

  (7) 0x401550 LDR	D25, [X15, X9,LSL #3]

  (7) 0x401554 LDR	D24, [X13, X9,LSL #3]

  (7) 0x401558 FADD	D3, D25, D24

  (7) 0x40155c STR	D3, [X14, X9,LSL #3]

  (7) 0x401560 LDR	D23, [X15, X8,LSL #3]

  (7) 0x401564 LDR	D22, [X13, X8,LSL #3]

  (7) 0x401568 FADD	D4, D23, D22

  (7) 0x40156c STR	D4, [X14, X8,LSL #3]

  (7) 0x401570 LDR	D21, [X15, X18,LSL #3]

  (7) 0x401574 LDR	D20, [X13, X18,LSL #3]

  (7) 0x401578 FADD	D5, D21, D20

  (7) 0x40157c STR	D5, [X14, X18,LSL #3]

  (7) 0x401580 LDR	D19, [X15, X30,LSL #3]

  (7) 0x401584 LDR	D18, [X13, X30,LSL #3]

  (7) 0x401588 FADD	D6, D19, D18

  (7) 0x40158c STR	D6, [X14, X30,LSL #3]

  (7) 0x401590 CMP	X12, X16

  (7) 0x401594 B.NE	4014f0

0x401598 LDR	X19, [SP, #16]

0x40159c LDP	X29, X30, [SP], #32

0x4015a0 RET

0x4015a4 ADD	X2, X2, #1

0x4015a8 MOVZ	X3, #0

0x4015ac B	401400

/home/eoseret/Tools/QaaS/qaas_runs/175-492-1446/intel/stream/build/stream/src/stream.c: 346 - 348

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

346: #pragma omp parallel for

347: 	for (j=0; j<STREAM_ARRAY_SIZE; j++)

348: 	    c[j] = a[j]+b[j];

Coverage (%)	Name	Source Location	Module
►100.00+	gomp_thread_start	team.c:130	libgomp.so.1.0.0
○	start_thread		libc.so.6
○	thread_start		libc.so.6

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	1.23
CQA speedup if FP arith vectorized	1.10
CQA speedup if fully vectorized	1.54
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.02
Bottlenecks	P12,
Function	main._omp_fn.6
Source	stream.c:346-348
Source loop unroll info	NA
Source loop unroll confidence level	NA
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	8.83
CQA cycles if no scalar integer	7.17
CQA cycles if FP arith vectorized	8.00
CQA cycles if fully vectorized	5.75
Front-end cycles	8.25
P0 cycles	5.50
P1 cycles	5.50
P2 cycles	3.75
P3 cycles	3.75
P4 cycles	3.67
P5 cycles	3.67
P6 cycles	3.58
P7 cycles	3.58
P8 cycles	3.50
P9 cycles	3.50
P10 cycles	3.50
P11 cycles	3.50
P12 cycles	8.83
P13 cycles	8.50
P14 cycles	8.67
P15 cycles	0.00
P16 cycles	0.00
DIV/SQRT cycles	0.00
Inter-iter dependencies cycles	NA
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	66.00
Nb uops	66.00
Nb loads	NA
Nb stores	7.00
Nb stack references	2.00
FLOP/cycle	0.79
Nb FLOP add-sub	7.00
Nb FLOP mul	0.00
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	25.36
Bytes prefetched	0.00
Bytes loaded	168.00
Bytes stored	56.00
Stride 0	NA
Stride 1	NA
Stride n	NA
Stride unknown	NA
Stride indirect	NA
Vectorization ratio all	0.00
Vectorization ratio load	0.00
Vectorization ratio store	0.00
Vectorization ratio mul	NA
Vectorization ratio add_sub	0.00
Vectorization ratio fma	NA
Vectorization ratio div_sqrt	NA
Vectorization ratio other	0.00
Vector-efficiency ratio all	51.19
Vector-efficiency ratio load	53.13
Vector-efficiency ratio store	50.00
Vector-efficiency ratio mul	NA
Vector-efficiency ratio add_sub	50.00
Vector-efficiency ratio fma	NA
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	50.00

Metric	Value
CQA speedup if no scalar integer	1.23
CQA speedup if FP arith vectorized	1.10
CQA speedup if fully vectorized	1.54
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.02
Bottlenecks	P12,
Function	main._omp_fn.6
Source	stream.c:346-348
Source loop unroll info	NA
Source loop unroll confidence level	NA
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	8.83
CQA cycles if no scalar integer	7.17
CQA cycles if FP arith vectorized	8.00
CQA cycles if fully vectorized	5.75
Front-end cycles	8.25
P0 cycles	5.50
P1 cycles	5.50
P2 cycles	3.75
P3 cycles	3.75
P4 cycles	3.67
P5 cycles	3.67
P6 cycles	3.58
P7 cycles	3.58
P8 cycles	3.50
P9 cycles	3.50
P10 cycles	3.50
P11 cycles	3.50
P12 cycles	8.83
P13 cycles	8.50
P14 cycles	8.67
P15 cycles	0.00
P16 cycles	0.00
DIV/SQRT cycles	0.00
Inter-iter dependencies cycles	NA
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	66.00
Nb uops	66.00
Nb loads	NA
Nb stores	7.00
Nb stack references	2.00
FLOP/cycle	0.79
Nb FLOP add-sub	7.00
Nb FLOP mul	0.00
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	25.36
Bytes prefetched	0.00
Bytes loaded	168.00
Bytes stored	56.00
Stride 0	NA
Stride 1	NA
Stride n	NA
Stride unknown	NA
Stride indirect	NA
Vectorization ratio all	0.00
Vectorization ratio load	0.00
Vectorization ratio store	0.00
Vectorization ratio mul	NA
Vectorization ratio add_sub	0.00
Vectorization ratio fma	NA
Vectorization ratio div_sqrt	NA
Vectorization ratio other	0.00
Vector-efficiency ratio all	51.19
Vector-efficiency ratio load	53.13
Vector-efficiency ratio store	50.00
Vector-efficiency ratio mul	NA
Vector-efficiency ratio add_sub	50.00
Vector-efficiency ratio fma	NA
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	50.00

Path /

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

Function	main._omp_fn.6
Source file and lines	stream.c:346-348
Module	exec

The loop is defined in /home/eoseret/Tools/QaaS/qaas_runs/175-492-1446/intel/stream/build/stream/src/stream.c:346-348.

Warnings:

Non-innermost loop: analyzing only self part (ignoring child loops).
Ignoring paths for analysis
Failed to get the number of paths
RecMII not computed since number of paths is unknown or > max_paths
Streams not analyzed since number of paths is unknown or > max_paths

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 8.83 to 7.17 cycles (1.23x speedup).

Workaround

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

Vectorization

Your loop is not vectorized. Only 51% of vector register length is used (average across all VPU instructions). By vectorizing your loop, you can lower the cost of an iteration from 8.83 to 5.75 cycles (1.54x speedup).

Details

All VPU instructions are used in scalar version (process only one data element in vector registers). Since your execution units are vector units, only a 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)

No data for this section

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

The binary loop is composed of 7 FP arithmetical operations:

7: addition or subtraction

The binary loop is loading 168 bytes. The binary loop is storing 56 bytes.

Arithmetic intensity

Arithmetic intensity is 0.03 FP operations per loaded or stored byte.

General properties

nb instructions	66
nb uops	66
loop length	264
used w registers	0
used x registers	15
used b registers	0
used h registers	0
used s registers	1
used d registers	21
used q registers	0
used v registers	0
used z registers	0
nb stack references	2

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	8.25 cycles
front end	8.25 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15	P16
uops	5.50	5.50	3.75	3.75	3.67	3.67	3.58	3.58	3.50	3.50	3.50	3.50	8.83	8.50	8.67	0.00	0.00
cycles	5.50	5.50	3.75	3.75	3.67	3.67	3.58	3.58	3.50	3.50	3.50	3.50	8.83	8.50	8.67	0.00	0.00

Execution ports to units layout:

P0: BRU
P1: BRU
P2: ALU
P3: ALU
P4: ALU
P5: ALU
P6: ALU
P7: ALU
P8 (128 bits): VPU, FP store data, ALU, DIV/SQRT
P9 (128 bits): VPU, ALU, FP store data
P10 (128 bits): VPU, ALU, DIV/SQRT
P11 (128 bits): ALU, VPU
P12 (256 bits): store address, load
P13 (256 bits): store address, load
P14 (256 bits): load
P15 (64 bits): store data
P16 (64 bits): store data

Cycles executing div or sqrt instructions

Cycles summary

Front-end	8.25
Dispatch	8.83
Overall L1	8.83

Vectorization ratios

INT

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

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

INT+FP

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

Vector efficiency ratios

INT

all	51%
load	53%
store	50%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	50%
fma	NA (no fma vectorizable/vectorized instructions)
other	50%

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

INT+FP

all	51%
load	53%
store	50%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	50%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	50%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 8.83 cycles. At this rate:

19% of peak load performance is reached (19.02 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
13% of peak store performance is reached (6.34 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the loop is: 4015a4

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput	Vectorization
MADD X16, X2, X0, X3	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X12, X2, X16	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	scal (50.0%)
CMP X16, X12	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.GE 401598 <main._omp_fn.6+0x1c8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADRP X4, 401410 <main._omp_fn.6+0x40>	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
SUB X5, X12, X16	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	scal (50.0%)
LDR X6, [X4, #1456]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33	N/A
ANDS X7, X5, #4224	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
LDR X15, [X6, #16]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33	N/A
LDP X13, X14, [X6]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.50	N/A
B.EQ 4014f0 <main._omp_fn.6+0x120>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X7, #1	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.EQ 4014d4 <main._omp_fn.6+0x104>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X7, #2	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.EQ 4014c0 <main._omp_fn.6+0xf0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X7, #3	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.EQ 4014ac <main._omp_fn.6+0xdc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X7, #4	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.EQ 401498 <main._omp_fn.6+0xc8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X7, #5	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.EQ 401484 <main._omp_fn.6+0xb4>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X7, #6	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.EQ 401470 <main._omp_fn.6+0xa0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR D31, [X15, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
LDR D30, [X13, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
FADD D0, D31, D30	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (50.0%)
STR D0, [X14, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (50.0%)
ADD X16, X16, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
LDR D29, [X15, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
LDR D28, [X13, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
FADD D1, D29, D28	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (50.0%)
STR D1, [X14, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (50.0%)
ADD X16, X16, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
LDR D27, [X15, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
LDR D26, [X13, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
FADD D2, D27, D26	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (50.0%)
STR D2, [X14, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (50.0%)
ADD X16, X16, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
LDR D25, [X15, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
LDR D24, [X13, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
FADD D3, D25, D24	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (50.0%)
STR D3, [X14, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (50.0%)
ADD X16, X16, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
LDR D23, [X15, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
LDR D22, [X13, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
FADD D4, D23, D22	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (50.0%)
STR D4, [X14, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (50.0%)
ADD X16, X16, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
LDR D21, [X15, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
LDR D20, [X13, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
FADD D5, D21, D20	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (50.0%)
STR D5, [X14, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (50.0%)
ADD X16, X16, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
LDR D19, [X15, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
LDR D18, [X13, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
FADD D6, D19, D18	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (50.0%)
STR D6, [X14, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (50.0%)
ADD X16, X16, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
CMP X12, X16	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.EQ 401598 <main._omp_fn.6+0x1c8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR X19, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33	scal (50.0%)
LDP X29, X30, [SP], #32	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (100.0%)
RET	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADD X2, X2, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
MOVZ X3, #0	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	scal (50.0%)
B 401400 <main._omp_fn.6+0x30>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A

Function	main._omp_fn.6
Source file and lines	stream.c:346-348
Module	exec

The loop is defined in /home/eoseret/Tools/QaaS/qaas_runs/175-492-1446/intel/stream/build/stream/src/stream.c:346-348.

Warnings:

Non-innermost loop: analyzing only self part (ignoring child loops).
Ignoring paths for analysis
Failed to get the number of paths
RecMII not computed since number of paths is unknown or > max_paths
Streams not analyzed since number of paths is unknown or > max_paths

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 8.83 to 7.17 cycles (1.23x speedup).

Workaround

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

Vectorization

Details

All VPU instructions are used in scalar version (process only one data element in vector registers). Since your execution units are vector units, only a 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)

No data for this section

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

The binary loop is composed of 7 FP arithmetical operations:

7: addition or subtraction

The binary loop is loading 168 bytes. The binary loop is storing 56 bytes.

Arithmetic intensity

Arithmetic intensity is 0.03 FP operations per loaded or stored byte.

General properties

nb instructions	66
nb uops	66
loop length	264
used w registers	0
used x registers	15
used b registers	0
used h registers	0
used s registers	1
used d registers	21
used q registers	0
used v registers	0
used z registers	0
nb stack references	2

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	8.25 cycles
front end	8.25 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15	P16
uops	5.50	5.50	3.75	3.75	3.67	3.67	3.58	3.58	3.50	3.50	3.50	3.50	8.83	8.50	8.67	0.00	0.00
cycles	5.50	5.50	3.75	3.75	3.67	3.67	3.58	3.58	3.50	3.50	3.50	3.50	8.83	8.50	8.67	0.00	0.00

Execution ports to units layout:

P0: BRU
P1: BRU
P2: ALU
P3: ALU
P4: ALU
P5: ALU
P6: ALU
P7: ALU
P8 (128 bits): VPU, FP store data, ALU, DIV/SQRT
P9 (128 bits): VPU, ALU, FP store data
P10 (128 bits): VPU, ALU, DIV/SQRT
P11 (128 bits): ALU, VPU
P12 (256 bits): store address, load
P13 (256 bits): store address, load
P14 (256 bits): load
P15 (64 bits): store data
P16 (64 bits): store data

Cycles executing div or sqrt instructions

Cycles summary

Front-end	8.25
Dispatch	8.83
Overall L1	8.83

Vectorization ratios

INT

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

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

INT+FP

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

Vector efficiency ratios

INT

all	51%
load	53%
store	50%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	50%
fma	NA (no fma vectorizable/vectorized instructions)
other	50%

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

INT+FP

all	51%
load	53%
store	50%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	50%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	50%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 8.83 cycles. At this rate:

19% of peak load performance is reached (19.02 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
13% of peak store performance is reached (6.34 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the loop is: 4015a4

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput	Vectorization
MADD X16, X2, X0, X3	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X12, X2, X16	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	scal (50.0%)
CMP X16, X12	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.GE 401598 <main._omp_fn.6+0x1c8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADRP X4, 401410 <main._omp_fn.6+0x40>	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
SUB X5, X12, X16	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	scal (50.0%)
LDR X6, [X4, #1456]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33	N/A
ANDS X7, X5, #4224	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
LDR X15, [X6, #16]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33	N/A
LDP X13, X14, [X6]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.50	N/A
B.EQ 4014f0 <main._omp_fn.6+0x120>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X7, #1	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.EQ 4014d4 <main._omp_fn.6+0x104>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X7, #2	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.EQ 4014c0 <main._omp_fn.6+0xf0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X7, #3	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.EQ 4014ac <main._omp_fn.6+0xdc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X7, #4	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.EQ 401498 <main._omp_fn.6+0xc8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X7, #5	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.EQ 401484 <main._omp_fn.6+0xb4>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X7, #6	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.EQ 401470 <main._omp_fn.6+0xa0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR D31, [X15, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
LDR D30, [X13, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
FADD D0, D31, D30	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (50.0%)
STR D0, [X14, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (50.0%)
ADD X16, X16, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
LDR D29, [X15, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
LDR D28, [X13, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
FADD D1, D29, D28	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (50.0%)
STR D1, [X14, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (50.0%)
ADD X16, X16, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
LDR D27, [X15, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
LDR D26, [X13, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
FADD D2, D27, D26	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (50.0%)
STR D2, [X14, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (50.0%)
ADD X16, X16, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
LDR D25, [X15, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
LDR D24, [X13, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
FADD D3, D25, D24	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (50.0%)
STR D3, [X14, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (50.0%)
ADD X16, X16, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
LDR D23, [X15, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
LDR D22, [X13, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
FADD D4, D23, D22	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (50.0%)
STR D4, [X14, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (50.0%)
ADD X16, X16, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
LDR D21, [X15, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
LDR D20, [X13, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
FADD D5, D21, D20	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (50.0%)
STR D5, [X14, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (50.0%)
ADD X16, X16, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
LDR D19, [X15, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
LDR D18, [X13, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (50.0%)
FADD D6, D19, D18	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (50.0%)
STR D6, [X14, X16,LSL #3]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (50.0%)
ADD X16, X16, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
CMP X12, X16	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (50.0%)
B.EQ 401598 <main._omp_fn.6+0x1c8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR X19, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33	scal (50.0%)
LDP X29, X30, [SP], #32	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (100.0%)
RET	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADD X2, X2, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
MOVZ X3, #0	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	scal (50.0%)
B 401400 <main._omp_fn.6+0x30>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A

Report Configuration

Code clean check

Workaround

Vectorization

Details

Workaround

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

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