OV - exec - Loop 182

	Loop Id: 182	Module: exec	Source: advec_mom.cpp:86-88 [...]	Coverage: 1.10%

0x41dec0 ADD	X2, X14, X16

0x41dec4 ADD	W6, W1, W16

0x41dec8 ADD	W7, W0, W16

0x41decc ADD	X16, X16, #2

0x41ded0 ADD	X3, X2, #1

0x41ded4 SDIV	X2, X2, X25

0x41ded8 ADD	W4, W23, W2

0x41dedc SDIV	X3, X3, X25

0x41dee0 ADD	W21, W22, W2

0x41dee4 SBFM	X4, X4, #0, #31

0x41dee8 ADD	W5, W23, W3

0x41deec SBFM	X21, X21, #0, #31

0x41def0 MADD	X28, X9, X4, XZR

0x41def4 SBFM	X5, X5, #0, #31

0x41def8 MADD	X21, X9, X21, XZR

0x41defc MADD	X30, X9, X5, XZR

0x41df00 ADD	X28, X8, X28,LSL #3

0x41df04 ADD	X21, X8, X21,LSL #3

0x41df08 ADD	X30, X8, X30,LSL #3

0x41df0c MADD	W6, W18, W2, W6

0x41df10 LDR	D2, [X28, X6,SXTW #3]    [7]

0x41df14 LDR	D1, [X21, X6,SXTW #3]    [1]

0x41df18 MADD	W20, W18, W3, W7

0x41df1c SBFM	X26, X20, #61, #31

0x41df20 ADD	W3, W22, W3

0x41df24 SBFM	X3, X3, #0, #31

0x41df28 MADD	X3, X9, X3, XZR

0x41df2c ADD	X3, X8, X3,LSL #3

0x41df30 ADD	X15, X30, X26

0x41df34 ADD	X27, X3, X26

0x41df38 LD1	{V2.D[1]}, [X15]    [4]

0x41df3c LD1	{V1.D[1]}, [X27]    [5]

0x41df40 FADD	V1.2D, V2.2D, V1.2D

0x41df44 MADD	W15, W18, W2, W7

0x41df48 LDR	D3, [X21, X15,SXTW #3]    [1]

0x41df4c LDR	D4, [X28, X15,SXTW #3]    [7]

0x41df50 ADD	W2, W20, #1

0x41df54 SBFM	X2, X2, #61, #31

0x41df58 ADD	X3, X3, X2

0x41df5c ADD	X2, X30, X2

0x41df60 MADD	X15, X11, X4, XZR

0x41df64 ADD	X15, X10, X15,LSL #3

0x41df68 LD1	{V3.D[1]}, [X3]    [8]

0x41df6c LD1	{V4.D[1]}, [X2]    [6]

0x41df70 MADD	X2, X11, X5, XZR

0x41df74 ADD	X2, X10, X2,LSL #3

0x41df78 ADD	X2, X2, X26

0x41df7c FADD	V2.2D, V3.2D, V4.2D

0x41df80 FADD	V1.2D, V1.2D, V2.2D

0x41df84 FMUL	V1.2D, V1.2D, V0.2D

0x41df88 STR	D1, [X15, X6,SXTW #3]    [2]

0x41df8c ST1	{V1.D[1]}, [X2]    [3]

0x41df90 CMP	X17, X16

0x41df94 B.NE	41dec0

/home/hbollore/qaas/qaas-runs/174-169-1878/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_mom.cpp: 86 - 88

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

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

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

88:           node_flux(i, j) = 0.25 * (mass_flux_x(i + 0, j - 1) + mass_flux_x(i, j) + mass_flux_x(i + 1, j - 1) + mass_flux_x(i + 1, j + 0));

/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	3.30 - 12.00
CQA speedup if FP arith vectorized	1.00
CQA speedup if fully vectorized	1.10 - 1.00
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	3.64
Bottlenecks	P8, P10,
Function	advec_mom_kernel(int, int, int, int, 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::Buffer1D&, clover::Buffer1D&, int, int, int) [clone .omp_outlined.8]
Source	advec_mom.cpp:86-88,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	11.00 - 40.00
CQA cycles if no scalar integer	3.33
CQA cycles if FP arith vectorized	11.00 - 40.00
CQA cycles if fully vectorized	10.00 - 40.00
Front-end cycles	6.75
P0 cycles	0.50
P1 cycles	0.50
P2 cycles	5.67
P3 cycles	5.58
P4 cycles	5.67
P5 cycles	5.50
P6 cycles	11.00
P7 cycles	5.58
P8 cycles	2.50
P9 cycles	2.50
P10 cycles	2.50
P11 cycles	2.50
P12 cycles	3.33
P13 cycles	3.33
P14 cycles	3.33
P15 cycles	0.00
P16 cycles	0.00
DIV/SQRT cycles	10.00 - 40.00
Inter-iter dependencies cycles	1
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	54.00
Nb uops	54.00
Nb loads	NA
Nb stores	2.00
Nb stack references	0.00
FLOP/cycle	0.73 - 0.20
Nb FLOP add-sub	6.00
Nb FLOP mul	2.00
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	2.20 - 8.00
Bytes prefetched	0.00
Bytes loaded	64.00
Bytes stored	24.00
Stride 0	0.00
Stride 1	0.00
Stride n	0.00
Stride unknown	5.00
Stride indirect	3.00
Vectorization ratio all	17.39
Vectorization ratio load	0.00
Vectorization ratio store	0.00
Vectorization ratio mul	100.00
Vectorization ratio add_sub	42.86
Vectorization ratio fma	0.00
Vectorization ratio div_sqrt	NA
Vectorization ratio other	0.00
Vector-efficiency ratio all	60.87
Vector-efficiency ratio load	50.00
Vector-efficiency ratio store	50.00
Vector-efficiency ratio mul	100.00
Vector-efficiency ratio add_sub	60.71
Vector-efficiency ratio fma	37.50
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	100.00

Metric	Value
CQA speedup if no scalar integer	3.30 - 12.00
CQA speedup if FP arith vectorized	1.00
CQA speedup if fully vectorized	1.10 - 1.00
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	3.64
Bottlenecks	P8, P10,
Function	advec_mom_kernel(int, int, int, int, 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::Buffer1D&, clover::Buffer1D&, int, int, int) [clone .omp_outlined.8]
Source	advec_mom.cpp:86-88,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	11.00 - 40.00
CQA cycles if no scalar integer	3.33
CQA cycles if FP arith vectorized	11.00 - 40.00
CQA cycles if fully vectorized	10.00 - 40.00
Front-end cycles	6.75
P0 cycles	0.50
P1 cycles	0.50
P2 cycles	5.67
P3 cycles	5.58
P4 cycles	5.67
P5 cycles	5.50
P6 cycles	11.00
P7 cycles	5.58
P8 cycles	2.50
P9 cycles	2.50
P10 cycles	2.50
P11 cycles	2.50
P12 cycles	3.33
P13 cycles	3.33
P14 cycles	3.33
P15 cycles	0.00
P16 cycles	0.00
DIV/SQRT cycles	10.00 - 40.00
Inter-iter dependencies cycles	1
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	54.00
Nb uops	54.00
Nb loads	NA
Nb stores	2.00
Nb stack references	0.00
FLOP/cycle	0.73 - 0.20
Nb FLOP add-sub	6.00
Nb FLOP mul	2.00
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	2.20 - 8.00
Bytes prefetched	0.00
Bytes loaded	64.00
Bytes stored	24.00
Stride 0	0.00
Stride 1	0.00
Stride n	0.00
Stride unknown	5.00
Stride indirect	3.00
Vectorization ratio all	17.39
Vectorization ratio load	0.00
Vectorization ratio store	0.00
Vectorization ratio mul	100.00
Vectorization ratio add_sub	42.86
Vectorization ratio fma	0.00
Vectorization ratio div_sqrt	NA
Vectorization ratio other	0.00
Vector-efficiency ratio all	60.87
Vector-efficiency ratio load	50.00
Vector-efficiency ratio store	50.00
Vector-efficiency ratio mul	100.00
Vector-efficiency ratio add_sub	60.71
Vector-efficiency ratio fma	37.50
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	100.00

Path /

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

Function	advec_mom_kernel(int, int, int, int, 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::Buffer1D&, clover::Buffer1D&, int, int, int) [clone .omp_outlined.8]
Source file and lines	advec_mom.cpp:86-88
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_mom.cpp:86-88
/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 40.00 to 3.33 cycles (12.00x 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 60% of vector register length is used (average across all VPU instructions).

Details

17% 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.
42% of VPU addition or subtraction instructions are used in vector version.
0% of VPU fused multiply-add 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.

FMA

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 8 FP arithmetical operations:

6: addition or subtraction
2: multiply

The binary loop is loading 64 bytes. The binary loop is storing 24 bytes.

Arithmetic intensity

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

General properties

nb instructions	54
loop length	216
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

6.75 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	5.67	5.58	5.67	5.50	11.00	5.58	2.50	2.50	2.50	2.50	3.33	3.33	3.33	0.00	0.00
cycles	0.50	0.50	5.67	5.58	5.67	5.50	11.00	5.58	2.50	2.50	2.50	2.50	3.33	3.33	3.33	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	10.00-40.00
Longest recurrence chain latency (RecMII)	1.00

Cycles summary

Front-end	6.75
Data deps.	1.00
Overall L1	11.00-40.00

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	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	17%
load	0%
store	0%
mul	100%
add-sub	42%
fma	0%
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	0%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 40.00 cycles. At this rate:

1% of peak load performance is reached (1.60 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
1% of peak store performance is reached (0.60 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the loop is: 41dec0

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput	Vectorization
ADD X2, X14, 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	N/A
ADD W6, W1, W16	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 W7, W0, W16	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%)
ADD X16, X16, #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 X3, 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
SDIV X2, X2, X25	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	5-20	5-20	N/A
ADD W4, W23, W2	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%)
SDIV X3, X3, X25	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	5-20	5-20	N/A
ADD W21, W22, W2	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 X4, X4, #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 W5, W23, W3	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 X21, X21, #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	N/A
MADD X28, X9, X4, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
SBFM X5, X5, #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 X21, X9, X21, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X30, X9, X5, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (50.0%)
ADD X28, X8, X28,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 X21, X8, X21,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 X30, X8, X30,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%)
MADD W6, W18, W2, W6	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
LDR D2, [X28, X6,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 D1, [X21, X6,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 W20, W18, W3, W7	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
SBFM X26, X20, #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 W3, W22, W3	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 X3, X3, #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	N/A
MADD X3, X9, X3, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X3, X8, 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 X15, X30, X26	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 X27, X3, X26	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 {V2.D[1]}, [X15]	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 {V1.D[1]}, [X27]	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 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%)
MADD W15, W18, W2, W7	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
LDR D3, [X21, X15,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, [X28, X15,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 W2, 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
SBFM X2, X2, #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	N/A
ADD X3, X3, X2	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 X2, X30, X2	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 X15, X11, X4, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X15, X10, X15,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 {V3.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%)
LD1 {V4.D[1]}, [X2]	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 X2, X11, X5, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X2, X10, X2,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 X2, X2, X26	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 V2.2D, V3.2D, 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%)
FADD V1.2D, V1.2D, 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%)
FMUL V1.2D, V1.2D, V0.2D	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%)
STR D1, [X15, X6,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]}, [X2]	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 X17, X16	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 41dec0 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii.omp_outlined.8+0x100>	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_mom_kernel(int, int, int, int, 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::Buffer1D&, clover::Buffer1D&, int, int, int) [clone .omp_outlined.8]
Source file and lines	advec_mom.cpp:86-88
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_mom.cpp:86-88
/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

Your loop is not vectorized. Only 60% of vector register length is used (average across all VPU instructions).

Details

17% 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.
42% of VPU addition or subtraction instructions are used in vector version.
0% of VPU fused multiply-add 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.

FMA

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 8 FP arithmetical operations:

6: addition or subtraction
2: multiply

The binary loop is loading 64 bytes. The binary loop is storing 24 bytes.

Arithmetic intensity

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

General properties

nb instructions	54
loop length	216
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

6.75 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	5.67	5.58	5.67	5.50	11.00	5.58	2.50	2.50	2.50	2.50	3.33	3.33	3.33	0.00	0.00
cycles	0.50	0.50	5.67	5.58	5.67	5.50	11.00	5.58	2.50	2.50	2.50	2.50	3.33	3.33	3.33	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	10.00-40.00
Longest recurrence chain latency (RecMII)	1.00

Cycles summary

Front-end	6.75
Data deps.	1.00
Overall L1	11.00-40.00

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	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	17%
load	0%
store	0%
mul	100%
add-sub	42%
fma	0%
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	0%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 40.00 cycles. At this rate:

1% of peak load performance is reached (1.60 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
1% of peak store performance is reached (0.60 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the loop is: 41dec0

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput	Vectorization
ADD X2, X14, 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	N/A
ADD W6, W1, W16	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 W7, W0, W16	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%)
ADD X16, X16, #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 X3, 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
SDIV X2, X2, X25	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	5-20	5-20	N/A
ADD W4, W23, W2	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%)
SDIV X3, X3, X25	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	5-20	5-20	N/A
ADD W21, W22, W2	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 X4, X4, #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 W5, W23, W3	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 X21, X21, #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	N/A
MADD X28, X9, X4, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
SBFM X5, X5, #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 X21, X9, X21, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X30, X9, X5, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (50.0%)
ADD X28, X8, X28,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 X21, X8, X21,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 X30, X8, X30,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%)
MADD W6, W18, W2, W6	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
LDR D2, [X28, X6,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 D1, [X21, X6,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 W20, W18, W3, W7	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
SBFM X26, X20, #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 W3, W22, W3	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 X3, X3, #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	N/A
MADD X3, X9, X3, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X3, X8, 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 X15, X30, X26	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 X27, X3, X26	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 {V2.D[1]}, [X15]	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 {V1.D[1]}, [X27]	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 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%)
MADD W15, W18, W2, W7	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
LDR D3, [X21, X15,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, [X28, X15,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 W2, 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
SBFM X2, X2, #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	N/A
ADD X3, X3, X2	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 X2, X30, X2	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 X15, X11, X4, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X15, X10, X15,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 {V3.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%)
LD1 {V4.D[1]}, [X2]	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 X2, X11, X5, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X2, X10, X2,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 X2, X2, X26	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 V2.2D, V3.2D, 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%)
FADD V1.2D, V1.2D, 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%)
FMUL V1.2D, V1.2D, V0.2D	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%)
STR D1, [X15, X6,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]}, [X2]	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 X17, X16	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 41dec0 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii.omp_outlined.8+0x100>	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

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

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