OV - K-Means scalability gcc-O3 100000000 - Loop 10

	Loop Id: 10	Module: kmeans-gcc-O3	Source: main.cpp:138-143	Coverage: NA%

0x4020b8 MADD	W1, W5, W1, W0

0x4020bc ADD	W5, W5, W1

0x4020c0 CMP	W1, W5

0x4020c4 B.GE	40211c

0x4020c8 LDR	X2, [X29, #96]

0x4020cc SBFM	X0, X1, #0, #31

0x4020d0 ADD	X1, X2, W1,SXTW #4

0x4020d4 HINT	#0

0x4020d8 HINT	#0

0x4020dc HINT	#0

  (15) 0x4020e0 LDR	W2, [X26, X0,LSL #2]

  (15) 0x4020e4 ADD	X0, X0, #1

  (15) 0x4020e8 LDP	D24, D26, [X1], #16

  (15) 0x4020ec LDR	D25, [X23, X2,SXTW #3]

  (15) 0x4020f0 SBFM	X4, X2, #62, #31

  (15) 0x4020f4 LDR	D27, [X21, X2,SXTW #3]

  (15) 0x4020f8 LDR	W3, [X27, X4]

  (15) 0x4020fc FADD	D24, D25, D24

  (15) 0x402100 FADD	D26, D27, D26

  (15) 0x402104 ADD	W3, W3, #1

  (15) 0x402108 STR	W3, [X27, X4]

  (15) 0x40210c STR	D24, [X23, X2,SXTW #3]

  (15) 0x402110 STR	D26, [X21, X2,SXTW #3]

  (15) 0x402114 CMP	W5, W0

  (15) 0x402118 B.GT	4020e0

0x40211c LDR	X2, [X25, #24]

0x402120 LDR	X26, [X25, #40]

0x402124 LDR	X19, [X25, #56]

0x402128 STR	X2, [X29, #104]

0x40212c BL	401760

0x402130 CNTD	X3, ALL

0x402134 LDR	X2, [X29, #104]

0x402138 ADD	X0, X3, #1

0x40213c CMP	X24, X0

0x402140 B.CC	4023d4

0x402144 ADDPL	X6, X24, #63

0x402148 MOVN	X0, #0

0x40214c PTRUE	P7.B, ALL

0x402150 ADD	X6, X6, #1

0x402154 ADD	X4, X23, #8

0x402158 ADD	X1, X26, #8

0x40215c HINT	#0

  (14) 0x402160 LD1D	{Z23.D}, P7/Z, [X1, X0,LSL #3]

  (14) 0x402164 LD1D	{Z22.D}, P7/Z, [X4, X0,LSL #3]

  (14) 0x402168 ORR	X5, XZR, X0

  (14) 0x40216c FADD	Z22.D, Z23.D, Z22.D

  (14) 0x402170 ST1D	{Z22.D}, P7, [X1, X0,LSL #3]

  (14) 0x402174 ADD	X0, X0, X3

  (14) 0x402178 CMP	X6, X0

  (14) 0x40217c B.HI	402160

0x402180 ADDPL	X1, X5, #1

0x402184 ADD	X1, X1, #1

0x402188 ADD	X23, X23, X1,LSL #3

0x40218c ADD	X26, X26, X1,LSL #3

0x402190 CMP	X24, X0

0x402194 B.EQ	4021bc

0x402198 SUB	X0, X26, X1,LSL #3

0x40219c SUB	X23, X23, X1,LSL #3

  (7) 0x4021a0 LDR	D21, [X0, X1,LSL #3]

  (7) 0x4021a4 LDR	D20, [X23, X1,LSL #3]

  (7) 0x4021a8 FADD	D20, D21, D20

  (7) 0x4021ac STR	D20, [X0, X1,LSL #3]

  (7) 0x4021b0 ADD	X1, X1, #1

  (7) 0x4021b4 CMP	X24, X1

  (7) 0x4021b8 B.CS	4021a0

0x4021bc CNTD	X3, ALL

0x4021c0 ADD	X0, X3, #1

0x4021c4 CMP	X22, X0

0x4021c8 B.CC	4023dc

0x4021cc ADDPL	X6, X22, #63

0x4021d0 MOVN	X0, #0

0x4021d4 PTRUE	P7.B, ALL

0x4021d8 ADD	X6, X6, #1

0x4021dc ADD	X4, X21, #8

0x4021e0 ADD	X1, X19, #8

  (13) 0x4021e4 LD1D	{Z19.D}, P7/Z, [X1, X0,LSL #3]

  (13) 0x4021e8 LD1D	{Z18.D}, P7/Z, [X4, X0,LSL #3]

  (13) 0x4021ec ORR	X5, XZR, X0

  (13) 0x4021f0 FADD	Z18.D, Z19.D, Z18.D

  (13) 0x4021f4 ST1D	{Z18.D}, P7, [X1, X0,LSL #3]

  (13) 0x4021f8 ADD	X0, X0, X3

  (13) 0x4021fc CMP	X6, X0

  (13) 0x402200 B.HI	4021e4

0x402204 ADDPL	X1, X5, #1

0x402208 ADD	X1, X1, #1

0x40220c ADD	X21, X21, X1,LSL #3

0x402210 ADD	X19, X19, X1,LSL #3

0x402214 CMP	X22, X0

0x402218 B.EQ	402240

0x40221c SUB	X0, X19, X1,LSL #3

0x402220 SUB	X21, X21, X1,LSL #3

  (8) 0x402224 LDR	D17, [X0, X1,LSL #3]

  (8) 0x402228 LDR	D16, [X21, X1,LSL #3]

  (8) 0x40222c FADD	D16, D17, D16

  (8) 0x402230 STR	D16, [X0, X1,LSL #3]

  (8) 0x402234 ADD	X1, X1, #1

  (8) 0x402238 CMP	X22, X1

  (8) 0x40223c B.CS	402224

0x402240 CNTW	X3, ALL

0x402244 ADD	X0, X3, #3

0x402248 CMP	X20, X0

0x40224c B.CC	4023e4

0x402250 ADDPL	X6, X20, #62

0x402254 MOVN	X0, #0

0x402258 PTRUE	P7.B, ALL

0x40225c ADD	X6, X6, #1

0x402260 ADD	X4, X27, #4

0x402264 ADD	X1, X2, #4

  (12) 0x402268 LD1W	{Z7.S}, P7/Z, [X1, X0,LSL #2]

  (12) 0x40226c LD1W	{Z6.S}, P7/Z, [X4, X0,LSL #2]

  (12) 0x402270 ORR	X5, XZR, X0

  (12) 0x402274 ADD	Z6.S, Z7.S, Z6.S

  (12) 0x402278 ST1W	{Z6.S}, P7, [X1, X0,LSL #2]

  (12) 0x40227c ADD	X0, X0, X3

  (12) 0x402280 CMP	X0, X6

  (12) 0x402284 B.CC	402268

0x402288 ADDPL	X1, X5, #2

0x40228c ADD	X1, X1, #1

0x402290 ADD	X5, X27, X1,LSL #2

0x402294 ADD	X3, X2, X1,LSL #2

0x402298 CMP	X20, X0

0x40229c B.EQ	402340

0x4022a0 CNTD	X6, ALL

0x4022a4 SUB	X0, X20, X1

0x4022a8 ADD	X4, X6, #1

0x4022ac SUB	X28, X28, X1

0x4022b0 CMP	X0, X4

0x4022b4 B.CC	40231c

0x4022b8 UBFM	X4, X1, #62, #61

0x4022bc SUB	X8, X28, X6

0x4022c0 PTRUE	P7.B, ALL

0x4022c4 ADD	X4, X4, #4

0x4022c8 MOVN	X0, #0

0x4022cc ADD	X2, X2, X4

0x4022d0 ADD	X4, X27, X4

0x4022d4 HINT	#0

0x4022d8 HINT	#0

0x4022dc HINT	#0

  (11) 0x4022e0 LD1W	{Z5.D}, P7/Z, [X2, X0,LSL #2]

  (11) 0x4022e4 LD1W	{Z4.D}, P7/Z, [X4, X0,LSL #2]

  (11) 0x4022e8 ORR	X7, XZR, X0

  (11) 0x4022ec ADD	Z4.S, Z5.S, Z4.S

  (11) 0x4022f0 ST1W	{Z4.D}, P7, [X2, X0,LSL #2]

  (11) 0x4022f4 ADD	X0, X0, X6

  (11) 0x4022f8 CMP	X8, X0

  (11) 0x4022fc B.HI	4022e0

0x402300 ADDPL	X0, X7, #1

0x402304 ADD	X0, X0, #1

0x402308 ADD	X5, X5, X0,LSL #2

0x40230c ADD	X1, X1, X0

0x402310 ADD	X3, X3, X0,LSL #2

0x402314 CMP	X28, X0

0x402318 B.EQ	402340

0x40231c SUB	X2, X3, X1,LSL #2

0x402320 SUB	X4, X5, X1,LSL #2

  (9) 0x402324 LDR	W0, [X2, X1,LSL #2]

  (9) 0x402328 LDR	W3, [X4, X1,LSL #2]

  (9) 0x40232c ADD	W0, W0, W3

  (9) 0x402330 STR	W0, [X2, X1,LSL #2]

  (9) 0x402334 ADD	X1, X1, #1

  (9) 0x402338 CMP	X20, X1

  (9) 0x40233c B.CS	402324

0x402340 BL	4017c0

0x402344 ADD	SP, X29, #0

0x402348 LDP	X19, X20, [SP, #16]

0x40234c LDP	X21, X22, [SP, #32]

0x402350 LDP	X23, X24, [SP, #48]

0x402354 LDP	X25, X26, [SP, #64]

0x402358 LDP	X27, X28, [SP, #80]

0x40235c LDP	X29, X30, [SP], #112

0x402360 RET

0x402364 ADD	W5, W5, #1

0x402368 MOVZ	W0, #0

0x40236c B	4020b8

0x4023d4 MOVZ	X1, #0

0x4023d8 B	402198

0x4023dc MOVZ	X1, #0

0x4023e0 B	40221c

0x4023e4 MOVZ	X1, #0

0x4023e8 ORR	X3, XZR, X2

0x4023ec ORR	X5, XZR, X27

0x4023f0 B	4022a0

/home/fmusial/KMEANS_Benchmarks/kmeans/main.cpp: 138 - 143

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

138:         #pragma omp parallel for reduction(+:sum_x[:k], sum_y[:k], count[:k])

139:         for (int i = 0; i < n; ++i) {

140:             int cid = assignment[i];

141:             count[cid] += 1;

142:             sum_x[cid] += points[i].x;

143:             sum_y[cid] += points[i].y;

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	5.06
CQA speedup if FP arith vectorized	1.00
CQA speedup if fully vectorized	2.27
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.51
Bottlenecks	P2, P3, P4, P5,
Function	k_means(int, point_t, point_t, int*, int, int) [clone ._omp_fn.1]
Source	main.cpp:138-138
Source loop unroll info	NA
Source loop unroll confidence level	NA
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	20.25
CQA cycles if no scalar integer	4.00
CQA cycles if FP arith vectorized	20.25
CQA cycles if fully vectorized	8.94
Front-end cycles	13.38
P0 cycles	8.00
P1 cycles	8.00
P2 cycles	20.25
P3 cycles	20.25
P4 cycles	20.25
P5 cycles	20.25
P6 cycles	0.00
P7 cycles	0.00
P8 cycles	0.00
P9 cycles	0.00
P10 cycles	4.17
P11 cycles	3.83
P12 cycles	4.00
P13 cycles	0.50
P14 cycles	0.50
DIV/SQRT cycles	0.00
Inter-iter dependencies cycles	NA
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	114.00
Nb uops	107.00
Nb loads	NA
Nb stores	1.00
Nb stack references	6.00
FLOP/cycle	0.00
Nb FLOP add-sub	0.00
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	7.11
Bytes prefetched	0.00
Bytes loaded	136.00
Bytes stored	8.00
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	0.00
Vectorization ratio div_sqrt	NA
Vectorization ratio other	0.00
Vector-efficiency ratio all	30.27
Vector-efficiency ratio load	36.11
Vector-efficiency ratio store	25.00
Vector-efficiency ratio mul	NA
Vector-efficiency ratio add_sub	25.00
Vector-efficiency ratio fma	12.50
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	37.50

Metric	Value
CQA speedup if no scalar integer	5.06
CQA speedup if FP arith vectorized	1.00
CQA speedup if fully vectorized	2.27
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.51
Bottlenecks	P2, P3, P4, P5,
Function	k_means(int, point_t, point_t, int*, int, int) [clone ._omp_fn.1]
Source	main.cpp:138-138
Source loop unroll info	NA
Source loop unroll confidence level	NA
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	20.25
CQA cycles if no scalar integer	4.00
CQA cycles if FP arith vectorized	20.25
CQA cycles if fully vectorized	8.94
Front-end cycles	13.38
P0 cycles	8.00
P1 cycles	8.00
P2 cycles	20.25
P3 cycles	20.25
P4 cycles	20.25
P5 cycles	20.25
P6 cycles	0.00
P7 cycles	0.00
P8 cycles	0.00
P9 cycles	0.00
P10 cycles	4.17
P11 cycles	3.83
P12 cycles	4.00
P13 cycles	0.50
P14 cycles	0.50
DIV/SQRT cycles	0.00
Inter-iter dependencies cycles	NA
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	114.00
Nb uops	107.00
Nb loads	NA
Nb stores	1.00
Nb stack references	6.00
FLOP/cycle	0.00
Nb FLOP add-sub	0.00
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	7.11
Bytes prefetched	0.00
Bytes loaded	136.00
Bytes stored	8.00
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	0.00
Vectorization ratio div_sqrt	NA
Vectorization ratio other	0.00
Vector-efficiency ratio all	30.27
Vector-efficiency ratio load	36.11
Vector-efficiency ratio store	25.00
Vector-efficiency ratio mul	NA
Vector-efficiency ratio add_sub	25.00
Vector-efficiency ratio fma	12.50
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	37.50

Path /

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

Function	k_means(int, point_t, point_t, int*, int, int) [clone ._omp_fn.1]
Source file and lines	main.cpp:138-143
Module	kmeans-gcc-O3

The loop is defined in /home/fmusial/KMEANS_Benchmarks/kmeans/main.cpp:138-143.

Analyzed code is defined in /home/fmusial/KMEANS_Benchmarks/kmeans/main.cpp:138.

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 20.25 to 4.00 cycles (5.06x 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 30% of vector register length is used (average across all VPU instructions). By vectorizing your loop, you can lower the cost of an iteration from 20.25 to 8.94 cycles (2.27x 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:
- recompile with fassociative-math (included in Ofast or ffast-math) to extend loop vectorization to FP reductions.
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)

Expensive FP math instructions/calls

Detected performance impact from expensive FP math instructions/calls. By removing/reexpressing them, you can lower the cost of an iteration from 20.25 to 1.00 cycles (20.25x speedup).

CALL instructions

Detected function call instructions.

Details

Calling (and then returning from) a function prevents many compiler optimizations (like vectorization), breaks control flow (which reduces pipeline performance) and executes extra instructions to save/restore the registers used inside it, which is very expensive (dozens of cycles). Consider to inline small functions.

unknown: 2 occurrences►
- /home/fmusial/KMEANS_Benchmarks/kmeans/main.cpp:138
- /home/fmusial/KMEANS_Benchmarks/kmeans/main.cpp:138

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

The binary loop does not contain any FP arithmetical operations. The binary loop is loading 136 bytes. The binary loop is storing 8 bytes.

General properties

nb instructions	114
nb uops	107
loop length	456
used w registers	3
used x registers	22
used b registers	0
used h registers	0
used s registers	1
used d registers	0
used q registers	0
used v registers	0
used z registers	0
nb stack references	6

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	13.38 cycles
front end	13.38 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	8.00	8.00	20.25	20.25	20.25	20.25	0.00	0.00	0.00	0.00	4.17	3.83	4.00	0.50	0.50
cycles	8.00	8.00	20.25	20.25	20.25	20.25	0.00	0.00	0.00	0.00	4.17	3.83	4.00	0.50	0.50

Execution ports to units layout:

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

Cycles executing div or sqrt instructions

Cycles summary

Front-end	13.38
Dispatch	20.25
Overall L1	20.25

Vectorization ratios

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

Vector efficiency ratios

all	30%
load	36%
store	25%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	25%
fma	12%
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	37%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 20.25 cycles. At this rate:

6% of peak load performance is reached (6.72 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
0% of peak store performance is reached (0.40 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: 402364

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput	Vectorization
MADD W1, W5, W1, W0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (12.5%)
ADD W5, W5, W1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
CMP W1, W5	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
B.GE 40211c <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x2f8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR X2, [X29, #96]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
SBFM X0, X1, #0, #31	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (100.0%)
ADD X1, X2, W1,SXTW #4	1	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	0	0	2	0.50	scal (25.0%)
HINT #0																			N/A
HINT #0																			N/A
HINT #0																			N/A
LDR X2, [X25, #24]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
LDR X26, [X25, #40]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
LDR X19, [X25, #56]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
STR X2, [X29, #104]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50	scal (25.0%)
BL 401760 <@plt_start@+0xe0>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CNTD X3, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
LDR X2, [X29, #104]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
ADD X0, X3, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X24, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.CC 4023d4 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x5b0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADDPL X6, X24, #63	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
MOVN X0, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
PTRUE P7.B, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (100.0%)
ADD X6, X6, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X4, X23, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X1, X26, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
HINT #0																			N/A
ADDPL X1, X5, #1	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
ADD X1, X1, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X23, X23, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X26, X26, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X24, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4021bc <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x398>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB X0, X26, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
SUB X23, X23, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CNTD X3, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
ADD X0, X3, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X22, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.CC 4023dc <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x5b8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADDPL X6, X22, #63	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
MOVN X0, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
PTRUE P7.B, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (100.0%)
ADD X6, X6, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X4, X21, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X1, X19, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADDPL X1, X5, #1	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
ADD X1, X1, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X21, X21, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X19, X19, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X22, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 402240 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x41c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB X0, X19, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
SUB X21, X21, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CNTW X3, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
ADD X0, X3, #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X20, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.CC 4023e4 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x5c0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADDPL X6, X20, #62	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
MOVN X0, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
PTRUE P7.B, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (100.0%)
ADD X6, X6, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X4, X27, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X1, X2, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADDPL X1, X5, #2	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
ADD X1, X1, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X5, X27, X1,LSL #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
ADD X3, X2, X1,LSL #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X20, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 402340 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x51c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CNTD X6, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
SUB X0, X20, X1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X4, X6, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
SUB X28, X28, X1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X0, X4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.CC 40231c <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x4f8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
UBFM X4, X1, #62, #61	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
SUB X8, X28, X6	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
PTRUE P7.B, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (100.0%)
ADD X4, X4, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
MOVN X0, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X2, X2, X4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X4, X27, X4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
HINT #0																			N/A
HINT #0																			N/A
HINT #0																			N/A
ADDPL X0, X7, #1	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
ADD X0, X0, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X5, X5, X0,LSL #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
ADD X1, X1, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X3, X3, X0,LSL #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X28, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 402340 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x51c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB X2, X3, X1,LSL #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
SUB X4, X5, X1,LSL #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
BL 4017c0 <@plt_start@+0x140>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADD SP, X29, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
LDP X19, X20, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
LDP X21, X22, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
LDP X23, X24, [SP, #48]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
LDP X25, X26, [SP, #64]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	N/A
LDP X27, X28, [SP, #80]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
LDP X29, X30, [SP], #112	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	N/A
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 W5, W5, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
MOVZ W0, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
B 4020b8 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x294>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
MOVZ X1, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
B 402198 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x374>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
MOVZ X1, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
B 40221c <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x3f8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
MOVZ X1, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ORR X3, XZR, X2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ORR X5, XZR, X27	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
B 4022a0 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x47c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A

Function	k_means(int, point_t, point_t, int*, int, int) [clone ._omp_fn.1]
Source file and lines	main.cpp:138-143
Module	kmeans-gcc-O3

The loop is defined in /home/fmusial/KMEANS_Benchmarks/kmeans/main.cpp:138-143.

Analyzed code is defined in /home/fmusial/KMEANS_Benchmarks/kmeans/main.cpp:138.

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 20.25 to 4.00 cycles (5.06x 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:
- recompile with fassociative-math (included in Ofast or ffast-math) to extend loop vectorization to FP reductions.
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)

Expensive FP math instructions/calls

Detected performance impact from expensive FP math instructions/calls. By removing/reexpressing them, you can lower the cost of an iteration from 20.25 to 1.00 cycles (20.25x speedup).

CALL instructions

Detected function call instructions.

Details

unknown: 2 occurrences►
- /home/fmusial/KMEANS_Benchmarks/kmeans/main.cpp:138
- /home/fmusial/KMEANS_Benchmarks/kmeans/main.cpp:138

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

The binary loop does not contain any FP arithmetical operations. The binary loop is loading 136 bytes. The binary loop is storing 8 bytes.

General properties

nb instructions	114
nb uops	107
loop length	456
used w registers	3
used x registers	22
used b registers	0
used h registers	0
used s registers	1
used d registers	0
used q registers	0
used v registers	0
used z registers	0
nb stack references	6

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	13.38 cycles
front end	13.38 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	8.00	8.00	20.25	20.25	20.25	20.25	0.00	0.00	0.00	0.00	4.17	3.83	4.00	0.50	0.50
cycles	8.00	8.00	20.25	20.25	20.25	20.25	0.00	0.00	0.00	0.00	4.17	3.83	4.00	0.50	0.50

Execution ports to units layout:

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

Cycles executing div or sqrt instructions

Cycles summary

Front-end	13.38
Dispatch	20.25
Overall L1	20.25

Vectorization ratios

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

Vector efficiency ratios

all	30%
load	36%
store	25%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	25%
fma	12%
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	37%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 20.25 cycles. At this rate:

6% of peak load performance is reached (6.72 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
0% of peak store performance is reached (0.40 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: 402364

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput	Vectorization
MADD W1, W5, W1, W0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (12.5%)
ADD W5, W5, W1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
CMP W1, W5	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
B.GE 40211c <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x2f8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR X2, [X29, #96]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
SBFM X0, X1, #0, #31	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (100.0%)
ADD X1, X2, W1,SXTW #4	1	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	0	0	2	0.50	scal (25.0%)
HINT #0																			N/A
HINT #0																			N/A
HINT #0																			N/A
LDR X2, [X25, #24]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
LDR X26, [X25, #40]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
LDR X19, [X25, #56]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
STR X2, [X29, #104]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50	scal (25.0%)
BL 401760 <@plt_start@+0xe0>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CNTD X3, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
LDR X2, [X29, #104]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
ADD X0, X3, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X24, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.CC 4023d4 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x5b0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADDPL X6, X24, #63	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
MOVN X0, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
PTRUE P7.B, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (100.0%)
ADD X6, X6, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X4, X23, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X1, X26, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
HINT #0																			N/A
ADDPL X1, X5, #1	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
ADD X1, X1, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X23, X23, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X26, X26, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X24, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4021bc <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x398>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB X0, X26, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
SUB X23, X23, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CNTD X3, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
ADD X0, X3, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X22, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.CC 4023dc <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x5b8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADDPL X6, X22, #63	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
MOVN X0, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
PTRUE P7.B, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (100.0%)
ADD X6, X6, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X4, X21, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X1, X19, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADDPL X1, X5, #1	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
ADD X1, X1, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X21, X21, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X19, X19, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X22, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 402240 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x41c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB X0, X19, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
SUB X21, X21, X1,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CNTW X3, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
ADD X0, X3, #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X20, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.CC 4023e4 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x5c0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADDPL X6, X20, #62	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
MOVN X0, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
PTRUE P7.B, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (100.0%)
ADD X6, X6, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X4, X27, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X1, X2, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADDPL X1, X5, #2	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
ADD X1, X1, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X5, X27, X1,LSL #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
ADD X3, X2, X1,LSL #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X20, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 402340 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x51c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CNTD X6, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
SUB X0, X20, X1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X4, X6, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
SUB X28, X28, X1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X0, X4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.CC 40231c <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x4f8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
UBFM X4, X1, #62, #61	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
SUB X8, X28, X6	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
PTRUE P7.B, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (100.0%)
ADD X4, X4, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
MOVN X0, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X2, X2, X4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X4, X27, X4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
HINT #0																			N/A
HINT #0																			N/A
HINT #0																			N/A
ADDPL X0, X7, #1	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
ADD X0, X0, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X5, X5, X0,LSL #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
ADD X1, X1, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X3, X3, X0,LSL #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
CMP X28, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 402340 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x51c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB X2, X3, X1,LSL #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
SUB X4, X5, X1,LSL #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
BL 4017c0 <@plt_start@+0x140>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADD SP, X29, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
LDP X19, X20, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
LDP X21, X22, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
LDP X23, X24, [SP, #48]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
LDP X25, X26, [SP, #64]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	N/A
LDP X27, X28, [SP, #80]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
LDP X29, X30, [SP], #112	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	N/A
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 W5, W5, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
MOVZ W0, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
B 4020b8 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x294>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
MOVZ X1, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
B 402198 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x374>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
MOVZ X1, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
B 40221c <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x3f8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
MOVZ X1, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ORR X3, XZR, X2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ORR X5, XZR, X27	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
B 4022a0 <_Z7k_meansiP7point_tS0_Piii._omp_fn.1+0x47c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A

▶Scalability Runs Description

Run run_1_thread	Number processes: 1Number nodes: 1Run Command: <executable> kmeans/input/100000000.in 1000 100000000 50 25MPI Command: Dataset: Run Directory: /home/fmusial/KMEANS_BenchmarksOMP_PROC_BIND: closeOMP_NUM_THREADS: 1OMP_PLACES: cores
Run run_2_threads	Number processes: 1Number nodes: 1Run Command: <executable> kmeans/input/100000000.in 1000 100000000 50 25MPI Command: Dataset: Run Directory: /home/fmusial/KMEANS_BenchmarksOMP_PLACES: coresOMP_PROC_BIND: closeOMP_NUM_THREADS: 2
Run run_4_threads	Number processes: 1Number nodes: 1Run Command: <executable> kmeans/input/100000000.in 1000 100000000 50 25MPI Command: Dataset: Run Directory: /home/fmusial/KMEANS_BenchmarksOMP_PLACES: coresOMP_PROC_BIND: closeOMP_NUM_THREADS: 4
Run run_8_threads	Number processes: 1Number nodes: 1Run Command: <executable> kmeans/input/100000000.in 1000 100000000 50 25MPI Command: Dataset: Run Directory: /home/fmusial/KMEANS_BenchmarksOMP_PLACES: coresOMP_PROC_BIND: closeOMP_NUM_THREADS: 8
Run run_16_threads	Number processes: 1Number nodes: 1Run Command: <executable> kmeans/input/100000000.in 1000 100000000 50 25MPI Command: Dataset: Run Directory: /home/fmusial/KMEANS_BenchmarksOMP_PLACES: coresOMP_PROC_BIND: closeOMP_NUM_THREADS: 16
Run run_32_threads	Number processes: 1Number nodes: 1Run Command: <executable> kmeans/input/100000000.in 1000 100000000 50 25MPI Command: Dataset: Run Directory: /home/fmusial/KMEANS_BenchmarksOMP_PROC_BIND: closeOMP_PLACES: coresOMP_NUM_THREADS: 32
Run run_64_threads	Number processes: 1Number nodes: 1Run Command: <executable> kmeans/input/100000000.in 1000 100000000 50 25MPI Command: Dataset: Run Directory: /home/fmusial/KMEANS_BenchmarksOMP_PROC_BIND: closeOMP_PLACES: coresOMP_NUM_THREADS: 64

(run_1_thread) Efficiency	(run_1_thread) Potential Speed-Up (%)	(run_2_threads) Efficiency	(run_2_threads) Potential Speed-Up (%)	(run_4_threads) Efficiency	(run_4_threads) Potential Speed-Up (%)	(run_8_threads) Efficiency	(run_8_threads) Potential Speed-Up (%)	(run_16_threads) Efficiency	(run_16_threads) Potential Speed-Up (%)	(run_32_threads) Efficiency	(run_32_threads) Potential Speed-Up (%)	(run_64_threads) Efficiency	(run_64_threads) Potential Speed-Up (%)
												1	0

Speed-Up and Efficiency

▶Detailed Speed-Up and Efficiency

Run	Number of threads	Efficiency (ideal is 1)	Speedup	Ideal Speedup	Time (s)	Coverage (%)
run_1_thread
run_2_threads
run_4_threads
run_8_threads
run_16_threads
run_32_threads
run_64_threads	1	1	1	64	0.0050000003539026	0.0088912593200803

Report Configuration

Code clean check

Workaround

Vectorization

Details

Workaround

Expensive FP math instructions/calls

CALL instructions

Details

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

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

Expensive FP math instructions/calls

CALL instructions

Details

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

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code

▶Scalability Runs Description

▼Columns Filter

Speed-Up and Efficiency

▶Detailed Speed-Up and Efficiency