OV - exec - Loop 163

	Loop Id: 163	Module: exec	Source: advec_cell.cpp:118-125 [...]	Coverage: 3.73%

0x41bc20 LDUR	X0, [X29, #472]    [17]

0x41bc24 ADD	X4, X0, X5

0x41bc28 LDR	X0, [SP, #48]    [5]

0x41bc2c ADD	X19, X4, #1

0x41bc30 SDIV	X4, X4, X24

0x41bc34 ADD	W27, W0, W5

0x41bc38 LDUR	W0, [X29, #460]    [17]

0x41bc3c ADD	X5, X5, #2

0x41bc40 ADD	W25, W22, W4

0x41bc44 SDIV	X19, X19, X24

0x41bc48 ADD	W26, W22, W19

0x41bc4c SBFM	X28, X26, #0, #31

0x41bc50 MADD	X23, X11, X28, XZR

0x41bc54 MADD	X7, X17, X28, XZR

0x41bc58 MADD	X26, X9, X28, XZR

0x41bc5c ADD	X23, X10, X23,LSL #3

0x41bc60 ADD	X26, X8, X26,LSL #3

0x41bc64 MADD	W20, W0, W4, W27

0x41bc68 ADD	W21, W20, #1

0x41bc6c MADD	W4, W0, W19, W27

0x41bc70 SBFM	X27, X25, #0, #31

0x41bc74 ADD	W19, W4, #2

0x41bc78 SBFM	X30, X19, #61, #31

0x41bc7c MADD	X19, X11, X27, XZR

0x41bc80 MADD	X3, X17, X27, XZR

0x41bc84 MADD	X25, X9, X27, XZR

0x41bc88 ADD	X25, X8, X25,LSL #3

0x41bc8c LDR	D1, [X25, X21,SXTW #3]    [10]

0x41bc90 ADD	W4, W4, #3

0x41bc94 SBFM	X4, X4, #61, #31

0x41bc98 ADD	X19, X10, X19,LSL #3

0x41bc9c ADD	X26, X26, X30

0x41bca0 LDR	D0, [X19, X21,SXTW #3]    [16]

0x41bca4 ADD	X19, X23, X30

0x41bca8 MADD	X23, X13, X28, XZR

0x41bcac ADD	X23, X12, X23,LSL #3

0x41bcb0 ADD	X0, X23, X30

0x41bcb4 LD1	{V1.D[1]}, [X26]    [11]

0x41bcb8 LD1	{V0.D[1]}, [X19]    [15]

0x41bcbc MADD	X19, X13, X27, XZR

0x41bcc0 ADD	X6, X12, X19,LSL #3

0x41bcc4 ADD	W19, W20, #2

0x41bcc8 MADD	X20, X15, X27, XZR

0x41bccc ADD	X20, X14, X20,LSL #3

0x41bcd0 LDR	D2, [X6, X21,SXTW #3]    [3]

0x41bcd4 LDR	D4, [X20, X21,SXTW #3]    [6]

0x41bcd8 FMUL	V1.2D, V0.2D, V1.2D

0x41bcdc LD1	{V2.D[1]}, [X0]    [9]

0x41bce0 ADD	X0, X16, X3,LSL #3

0x41bce4 ADD	X3, X16, X7,LSL #3

0x41bce8 LDR	D3, [X0, X21,SXTW #3]    [8]

0x41bcec ADD	X7, X3, X30

0x41bcf0 ADD	X3, X3, X4

0x41bcf4 LD1	{V3.D[1]}, [X7]    [4]

0x41bcf8 MADD	X7, X15, X28, XZR

0x41bcfc ADD	X7, X14, X7,LSL #3

0x41bd00 ADD	X7, X7, X30

0x41bd04 LD1	{V4.D[1]}, [X7]    [12]

0x41bd08 FMLA	V3.2D, V4.2D, V1.2D

0x41bd0c LDR	D4, [X6, X19,SXTW #3]    [3]

0x41bd10 ADD	X6, X23, X4

0x41bd14 FADD	V1.2D, V2.2D, V1.2D

0x41bd18 LDR	D2, [X0, X19,SXTW #3]    [8]

0x41bd1c MADD	X0, X1, X27, XZR

0x41bd20 ADD	X0, X18, X0,LSL #3

0x41bd24 LD1	{V4.D[1]}, [X6]    [2]

0x41bd28 FSUB	V1.2D, V1.2D, V4.2D

0x41bd2c LDR	D4, [X0, X21,SXTW #3]    [7]

0x41bd30 LD1	{V2.D[1]}, [X3]    [14]

0x41bd34 MADD	X3, X1, X28, XZR

0x41bd38 ADD	X3, X18, X3,LSL #3

0x41bd3c ADD	X6, X3, X30

0x41bd40 ADD	X3, X3, X4

0x41bd44 FSUB	V2.2D, V3.2D, V2.2D

0x41bd48 LD1	{V4.D[1]}, [X6]    [1]

0x41bd4c FADD	V0.2D, V4.2D, V0.2D

0x41bd50 LDR	D4, [X0, X19,SXTW #3]    [7]

0x41bd54 LDUR	X0, [X29, #464]    [17]

0x41bd58 LD1	{V4.D[1]}, [X3]    [13]

0x41bd5c FSUB	V0.2D, V0.2D, V4.2D

0x41bd60 FDIV	V0.2D, V1.2D, V0.2D

0x41bd64 FDIV	V1.2D, V2.2D, V1.2D

0x41bd68 STR	D0, [X25, X21,SXTW #3]    [10]

0x41bd6c ST1	{V0.D[1]}, [X26]    [11]

0x41bd70 STR	D1, [X20, X21,SXTW #3]    [6]

0x41bd74 ST1	{V1.D[1]}, [X7]    [12]

0x41bd78 CMP	X0, X5

0x41bd7c B.NE	41bc20

/home/hbollore/qaas/qaas-runs/174-169-1878/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_cell.cpp: 118 - 125

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

118:     for (int j = (y_min + 1); j < (y_max + 2); j++) {

119:       for (int i = (x_min + 1); i < (x_max + 2); i++) {

120:         double pre_mass_s = density1(i, j) * pre_vol(i, j);

121:         double post_mass_s = pre_mass_s + mass_flux_x(i, j) - mass_flux_x(i + 1, j + 0);

122:         double post_ener_s = (energy1(i, j) * pre_mass_s + ener_flux(i, j) - ener_flux(i + 1, j + 0)) / post_mass_s;

123:         double advec_vol_s = pre_vol(i, j) + vol_flux_x(i, j) - vol_flux_x(i + 1, j + 0);

124:         density1(i, j) = post_mass_s / advec_vol_s;

125:         energy1(i, j) = post_ener_s;

/home/hbollore/qaas/qaas-runs/174-169-1878/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/context.h: 69 - 69

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

69:   T &operator()(size_t i, size_t j) const { return data[i + j * sizeX]; }

Coverage (%)	Name	Source Location	Module
○100.00	__kmp_invoke_microtask		libomp.so

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	2.04 - 3.86
CQA speedup if FP arith vectorized	1.00
CQA speedup if fully vectorized	1.17 - 1.23
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.06 - 3.37
Bottlenecks	P8, P10,
Function	advec_cell_kernel(int, int, int, int, int, int, clover::Buffer1D&, clover::Buffer1D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&) [clone .omp_outlined.6]
Source	advec_cell.cpp:118-125,context.h:69-69
Source loop unroll info	unrolled by 3
Source loop unroll confidence level	high
Unroll/vectorization loop type	NA
Unroll factor	3
CQA cycles	16.99 - 53.99
CQA cycles if no scalar integer	8.33 - 13.99
CQA cycles if FP arith vectorized	16.99 - 53.99
CQA cycles if fully vectorized	14.49 - 43.99
Front-end cycles	11.00
P0 cycles	0.50
P1 cycles	0.50
P2 cycles	7.17
P3 cycles	7.25
P4 cycles	7.17
P5 cycles	7.17
P6 cycles	16.00
P7 cycles	7.25
P8 cycles	5.50
P9 cycles	5.50
P10 cycles	5.50
P11 cycles	5.50
P12 cycles	8.67
P13 cycles	8.67
P14 cycles	8.67
P15 cycles	0.00
P16 cycles	0.00
DIV/SQRT cycles	16.99 - 53.99
Inter-iter dependencies cycles	1
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	88.00
Nb uops	88.00
Nb loads	NA
Nb stores	4.00
Nb stack references	0.00
FLOP/cycle	1.18 - 0.37
Nb FLOP add-sub	10.00
Nb FLOP mul	2.00
Nb FLOP fma	2.00
Nb FLOP div	4.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	4.08 - 12.95
Bytes prefetched	0.00
Bytes loaded	172.00
Bytes stored	48.00
Stride 0	2.00
Stride 1	0.00
Stride n	0.00
Stride unknown	9.00
Stride indirect	6.00
Vectorization ratio all	20.45
Vectorization ratio load	0.00
Vectorization ratio store	0.00
Vectorization ratio mul	100.00
Vectorization ratio add_sub	50.00
Vectorization ratio fma	25.00
Vectorization ratio div_sqrt	66.67
Vectorization ratio other	0.00
Vector-efficiency ratio all	63.07
Vector-efficiency ratio load	50.00
Vector-efficiency ratio store	50.00
Vector-efficiency ratio mul	100.00
Vector-efficiency ratio add_sub	70.00
Vector-efficiency ratio fma	56.25
Vector-efficiency ratio div_sqrt	83.33
Vector-efficiency ratio other	100.00

Metric	Value
CQA speedup if no scalar integer	2.04 - 3.86
CQA speedup if FP arith vectorized	1.00
CQA speedup if fully vectorized	1.17 - 1.23
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.06 - 3.37
Bottlenecks	P8, P10,
Function	advec_cell_kernel(int, int, int, int, int, int, clover::Buffer1D&, clover::Buffer1D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&) [clone .omp_outlined.6]
Source	advec_cell.cpp:118-125,context.h:69-69
Source loop unroll info	unrolled by 3
Source loop unroll confidence level	high
Unroll/vectorization loop type	NA
Unroll factor	3
CQA cycles	16.99 - 53.99
CQA cycles if no scalar integer	8.33 - 13.99
CQA cycles if FP arith vectorized	16.99 - 53.99
CQA cycles if fully vectorized	14.49 - 43.99
Front-end cycles	11.00
P0 cycles	0.50
P1 cycles	0.50
P2 cycles	7.17
P3 cycles	7.25
P4 cycles	7.17
P5 cycles	7.17
P6 cycles	16.00
P7 cycles	7.25
P8 cycles	5.50
P9 cycles	5.50
P10 cycles	5.50
P11 cycles	5.50
P12 cycles	8.67
P13 cycles	8.67
P14 cycles	8.67
P15 cycles	0.00
P16 cycles	0.00
DIV/SQRT cycles	16.99 - 53.99
Inter-iter dependencies cycles	1
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	88.00
Nb uops	88.00
Nb loads	NA
Nb stores	4.00
Nb stack references	0.00
FLOP/cycle	1.18 - 0.37
Nb FLOP add-sub	10.00
Nb FLOP mul	2.00
Nb FLOP fma	2.00
Nb FLOP div	4.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	4.08 - 12.95
Bytes prefetched	0.00
Bytes loaded	172.00
Bytes stored	48.00
Stride 0	2.00
Stride 1	0.00
Stride n	0.00
Stride unknown	9.00
Stride indirect	6.00
Vectorization ratio all	20.45
Vectorization ratio load	0.00
Vectorization ratio store	0.00
Vectorization ratio mul	100.00
Vectorization ratio add_sub	50.00
Vectorization ratio fma	25.00
Vectorization ratio div_sqrt	66.67
Vectorization ratio other	0.00
Vector-efficiency ratio all	63.07
Vector-efficiency ratio load	50.00
Vector-efficiency ratio store	50.00
Vector-efficiency ratio mul	100.00
Vector-efficiency ratio add_sub	70.00
Vector-efficiency ratio fma	56.25
Vector-efficiency ratio div_sqrt	83.33
Vector-efficiency ratio other	100.00

Path /

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

Function	advec_cell_kernel(int, int, int, int, int, int, clover::Buffer1D&, clover::Buffer1D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&) [clone .omp_outlined.6]
Source file and lines	advec_cell.cpp:118-125
Module	exec

The loop is defined in:

/home/hbollore/qaas/qaas-runs/174-169-1878/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_cell.cpp:118-125
/home/hbollore/qaas/qaas-runs/174-169-1878/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/context.h:69

The related source loop is unrolled by 3 (including vectorization).

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 53.99 to 13.99 cycles (3.86x 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 63% of vector register length is used (average across all VPU instructions). By fully vectorizing your loop, you can lower the cost of an iteration from 53.99 to 43.99 cycles (1.23x speedup).

Details

20% of VPU instructions are used in vector version (process two or more data elements in vector registers):

0% of VPU loads are used in vector version.
0% of VPU stores are used in vector version.
50% of VPU addition or subtraction instructions are used in vector version.
25% of VPU fused multiply-add instructions are used in vector version.
66% of VPU divide and square root instructions are used in vector version.
0% of VPU 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)

FMA

Detected 2 FMA (fused multiply-add) operations. Presence of both ADD/SUB and MUL operations.

Workaround

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).

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

The binary loop is composed of 20 FP arithmetical operations:

12: addition or subtraction (2 inside FMA instructions)
4: multiply (2 inside FMA instructions)
4: divide

The binary loop is loading 172 bytes. The binary loop is storing 48 bytes.

Arithmetic intensity

Arithmetic intensity is 0.09 FP operations per loaded or stored byte.

General properties

nb instructions	88
loop length	352
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

11.00 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15	P16
uops	0.50	0.50	7.17	7.25	7.17	7.17	16.00	7.25	5.50	5.50	5.50	5.50	8.67	8.67	8.67	0.00	0.00
cycles	0.50	0.50	7.17	7.25	7.17	7.17	16.00	7.25	5.50	5.50	5.50	5.50	8.67	8.67	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	16.99-53.99
Longest recurrence chain latency (RecMII)	1.00

Cycles summary

Front-end	11.00
Data deps.	1.00
Overall L1	16.99-53.99

Vectorization ratios

INT

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

all	100%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	100%
add-sub	100%
fma	100%
div/sqrt	100%
other	NA (no other vectorizable/vectorized instructions)

INT+FP

all	20%
load	0%
store	0%
mul	100%
add-sub	50%
fma	25%
div/sqrt	66%
other	0%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 53.99 cycles. At this rate:

3% of peak load performance is reached (3.19 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
1% of peak store performance is reached (0.89 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the loop is: 41bc20

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput	Vectorization
LDUR X0, [X29, #472]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33	N/A
ADD X4, X0, X5	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 X0, [SP, #48]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33	N/A
ADD X19, X4, #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
SDIV X4, X4, X24	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	5-20	5-20	scal (50.0%)
ADD W27, W0, W5	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 (25.0%)
LDUR W0, [X29, #460]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33	N/A
ADD X5, X5, #2	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
ADD W25, W22, W4	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
SDIV X19, X19, X24	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	5-20	5-20	N/A
ADD W26, W22, W19	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
SBFM X28, X26, #0, #31	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 (100.0%)
MADD X23, X11, X28, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (50.0%)
MADD X7, X17, X28, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X26, X9, X28, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X23, X10, X23,LSL #3	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%)
ADD X26, X8, X26,LSL #3	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
MADD W20, W0, W4, W27	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD W21, W20, #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
MADD W4, W0, W19, W27	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
SBFM X27, X25, #0, #31	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 (100.0%)
ADD W19, W4, #2	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
SBFM X30, X19, #61, #31	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 (100.0%)
MADD X19, X11, X27, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X3, X17, X27, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X25, X9, X27, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X25, X8, X25,LSL #3	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 D1, [X25, X21,SXTW #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%)
ADD W4, W4, #3	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 (25.0%)
SBFM X4, X4, #61, #31	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 (100.0%)
ADD X19, X10, X19,LSL #3	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
ADD X26, X26, X30	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 D0, [X19, X21,SXTW #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%)
ADD X19, X23, X30	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
MADD X23, X13, X28, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (50.0%)
ADD X23, X12, X23,LSL #3	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%)
ADD X0, X23, X30	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
LD1 {V1.D[1]}, [X26]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
LD1 {V0.D[1]}, [X19]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
MADD X19, X13, X27, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X6, X12, X19,LSL #3	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
ADD W19, W20, #2	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
MADD X20, X15, X27, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X20, X14, X20,LSL #3	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 D2, [X6, X21,SXTW #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 D4, [X20, X21,SXTW #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%)
FMUL V1.2D, V0.2D, V1.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	vect (100.0%)
LD1 {V2.D[1]}, [X0]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
ADD X0, X16, X3,LSL #3	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
ADD X3, X16, X7,LSL #3	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 D3, [X0, X21,SXTW #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%)
ADD X7, X3, X30	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
ADD X3, X3, X4	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
LD1 {V3.D[1]}, [X7]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
MADD X7, X15, X28, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X7, X14, X7,LSL #3	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
ADD X7, X7, X30	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
LD1 {V4.D[1]}, [X7]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
FMLA V3.2D, V4.2D, V1.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	vect (100.0%)
LDR D4, [X6, X19,SXTW #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%)
ADD X6, X23, X4	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
FADD V1.2D, V2.2D, V1.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	vect (100.0%)
LDR D2, [X0, X19,SXTW #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%)
MADD X0, X1, X27, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X0, X18, X0,LSL #3	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
LD1 {V4.D[1]}, [X6]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
FSUB V1.2D, V1.2D, V4.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	vect (100.0%)
LDR D4, [X0, X21,SXTW #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%)
LD1 {V2.D[1]}, [X3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
MADD X3, X1, X28, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X3, X18, X3,LSL #3	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
ADD X6, X3, X30	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
ADD X3, X3, X4	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
FSUB V2.2D, V3.2D, V2.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	vect (100.0%)
LD1 {V4.D[1]}, [X6]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
FADD V0.2D, V4.2D, V0.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	vect (100.0%)
LDR D4, [X0, X19,SXTW #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%)
LDUR X0, [X29, #464]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33	N/A
LD1 {V4.D[1]}, [X3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
FSUB V0.2D, V0.2D, V4.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	vect (100.0%)
FDIV V0.2D, V1.2D, V0.2D	1	0	0	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	3.50-6.99	vect (100.0%)
FDIV V1.2D, V2.2D, V1.2D	1	0	0	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	3.50-6.99	vect (100.0%)
STR D0, [X25, X21,SXTW #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%)
ST1 {V0.D[1]}, [X26]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	4	1	scal (50.0%)
STR D1, [X20, X21,SXTW #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%)
ST1 {V1.D[1]}, [X7]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	4	1	scal (50.0%)
CMP X0, X5	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	N/A
B.NE 41bc20 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_.omp_outlined.6+0x140>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A

Function	advec_cell_kernel(int, int, int, int, int, int, clover::Buffer1D&, clover::Buffer1D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&) [clone .omp_outlined.6]
Source file and lines	advec_cell.cpp:118-125
Module	exec

The loop is defined in:

/home/hbollore/qaas/qaas-runs/174-169-1878/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_cell.cpp:118-125
/home/hbollore/qaas/qaas-runs/174-169-1878/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/context.h:69

The related source loop is unrolled by 3 (including vectorization).

gain
potential
hint
expert

Code clean check

Workaround

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

Vectorization

Details

20% of VPU instructions are used in vector version (process two or more data elements in vector registers):

0% of VPU loads are used in vector version.
0% of VPU stores are used in vector version.
50% of VPU addition or subtraction instructions are used in vector version.
25% of VPU fused multiply-add instructions are used in vector version.
66% of VPU divide and square root instructions are used in vector version.
0% of VPU 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)

FMA

Detected 2 FMA (fused multiply-add) operations. Presence of both ADD/SUB and MUL operations.

Workaround

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

The binary loop is composed of 20 FP arithmetical operations:

12: addition or subtraction (2 inside FMA instructions)
4: multiply (2 inside FMA instructions)
4: divide

The binary loop is loading 172 bytes. The binary loop is storing 48 bytes.

Arithmetic intensity

Arithmetic intensity is 0.09 FP operations per loaded or stored byte.

General properties

nb instructions	88
loop length	352
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

11.00 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15	P16
uops	0.50	0.50	7.17	7.25	7.17	7.17	16.00	7.25	5.50	5.50	5.50	5.50	8.67	8.67	8.67	0.00	0.00
cycles	0.50	0.50	7.17	7.25	7.17	7.17	16.00	7.25	5.50	5.50	5.50	5.50	8.67	8.67	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	16.99-53.99
Longest recurrence chain latency (RecMII)	1.00

Cycles summary

Front-end	11.00
Data deps.	1.00
Overall L1	16.99-53.99

Vectorization ratios

INT

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

all	100%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	100%
add-sub	100%
fma	100%
div/sqrt	100%
other	NA (no other vectorizable/vectorized instructions)

INT+FP

all	20%
load	0%
store	0%
mul	100%
add-sub	50%
fma	25%
div/sqrt	66%
other	0%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 53.99 cycles. At this rate:

3% of peak load performance is reached (3.19 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
1% of peak store performance is reached (0.89 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the loop is: 41bc20

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput	Vectorization
LDUR X0, [X29, #472]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33	N/A
ADD X4, X0, X5	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 X0, [SP, #48]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33	N/A
ADD X19, X4, #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
SDIV X4, X4, X24	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	5-20	5-20	scal (50.0%)
ADD W27, W0, W5	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 (25.0%)
LDUR W0, [X29, #460]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33	N/A
ADD X5, X5, #2	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
ADD W25, W22, W4	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
SDIV X19, X19, X24	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	5-20	5-20	N/A
ADD W26, W22, W19	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
SBFM X28, X26, #0, #31	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 (100.0%)
MADD X23, X11, X28, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (50.0%)
MADD X7, X17, X28, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X26, X9, X28, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X23, X10, X23,LSL #3	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%)
ADD X26, X8, X26,LSL #3	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
MADD W20, W0, W4, W27	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD W21, W20, #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
MADD W4, W0, W19, W27	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
SBFM X27, X25, #0, #31	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 (100.0%)
ADD W19, W4, #2	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
SBFM X30, X19, #61, #31	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 (100.0%)
MADD X19, X11, X27, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X3, X17, X27, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X25, X9, X27, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X25, X8, X25,LSL #3	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 D1, [X25, X21,SXTW #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%)
ADD W4, W4, #3	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 (25.0%)
SBFM X4, X4, #61, #31	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 (100.0%)
ADD X19, X10, X19,LSL #3	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
ADD X26, X26, X30	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 D0, [X19, X21,SXTW #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%)
ADD X19, X23, X30	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
MADD X23, X13, X28, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (50.0%)
ADD X23, X12, X23,LSL #3	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%)
ADD X0, X23, X30	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
LD1 {V1.D[1]}, [X26]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
LD1 {V0.D[1]}, [X19]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
MADD X19, X13, X27, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X6, X12, X19,LSL #3	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
ADD W19, W20, #2	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
MADD X20, X15, X27, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X20, X14, X20,LSL #3	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 D2, [X6, X21,SXTW #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 D4, [X20, X21,SXTW #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%)
FMUL V1.2D, V0.2D, V1.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	vect (100.0%)
LD1 {V2.D[1]}, [X0]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
ADD X0, X16, X3,LSL #3	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
ADD X3, X16, X7,LSL #3	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 D3, [X0, X21,SXTW #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%)
ADD X7, X3, X30	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
ADD X3, X3, X4	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
LD1 {V3.D[1]}, [X7]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
MADD X7, X15, X28, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X7, X14, X7,LSL #3	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
ADD X7, X7, X30	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
LD1 {V4.D[1]}, [X7]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
FMLA V3.2D, V4.2D, V1.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	vect (100.0%)
LDR D4, [X6, X19,SXTW #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%)
ADD X6, X23, X4	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
FADD V1.2D, V2.2D, V1.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	vect (100.0%)
LDR D2, [X0, X19,SXTW #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%)
MADD X0, X1, X27, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X0, X18, X0,LSL #3	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
LD1 {V4.D[1]}, [X6]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
FSUB V1.2D, V1.2D, V4.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	vect (100.0%)
LDR D4, [X0, X21,SXTW #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%)
LD1 {V2.D[1]}, [X3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
MADD X3, X1, X28, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X3, X18, X3,LSL #3	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
ADD X6, X3, X30	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
ADD X3, X3, X4	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
FSUB V2.2D, V3.2D, V2.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	vect (100.0%)
LD1 {V4.D[1]}, [X6]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
FADD V0.2D, V4.2D, V0.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	vect (100.0%)
LDR D4, [X0, X19,SXTW #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%)
LDUR X0, [X29, #464]	1	0	0	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33	N/A
LD1 {V4.D[1]}, [X3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	8	0.33	scal (50.0%)
FSUB V0.2D, V0.2D, V4.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	vect (100.0%)
FDIV V0.2D, V1.2D, V0.2D	1	0	0	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	3.50-6.99	vect (100.0%)
FDIV V1.2D, V2.2D, V1.2D	1	0	0	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	3.50-6.99	vect (100.0%)
STR D0, [X25, X21,SXTW #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%)
ST1 {V0.D[1]}, [X26]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	4	1	scal (50.0%)
STR D1, [X20, X21,SXTW #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%)
ST1 {V1.D[1]}, [X7]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	4	1	scal (50.0%)
CMP X0, X5	1	0	0	0.25	0.25	0	0	0.25	0.25	0	0	0	0	0	0	0	1	0.33	N/A
B.NE 41bc20 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_.omp_outlined.6+0x140>	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

FMA

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

Cycles and memory resources usage

ASM code

Code clean check

Workaround

Vectorization

Details

Workaround

FMA

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

Cycles and memory resources usage

ASM code