OV - exec - Loop 190

	Loop Id: 190	Module: exec	Source: advec_mom.cpp:147-149 [...]	Coverage: 4.54%

0x41ea40 ADD	X5, X18, X1

0x41ea44 ADD	W21, W4, W1

0x41ea48 ADD	X1, X1, #2

0x41ea4c SDIV	X6, X5, X24

0x41ea50 ADD	X5, X5, #1

0x41ea54 ADD	W7, W22, W6

0x41ea58 SDIV	X5, X5, X24

0x41ea5c SBFM	X7, X7, #0, #31

0x41ea60 ADD	W20, W22, W5

0x41ea64 MADD	X28, X11, X7, XZR

0x41ea68 SBFM	X20, X20, #0, #31

0x41ea6c MADD	X26, X9, X7, XZR

0x41ea70 MADD	X30, X11, X20, XZR

0x41ea74 MADD	X27, X9, X20, XZR

0x41ea78 ADD	X28, X10, X28,LSL #3

0x41ea7c ADD	X26, X8, X26,LSL #3

0x41ea80 ADD	X30, X10, X30,LSL #3

0x41ea84 ADD	X27, X8, X27,LSL #3

0x41ea88 MADD	W6, W3, W6, W21

0x41ea8c ADD	W25, W6, #1

0x41ea90 LDR	D1, [X28, X25,SXTW #3]    [9]

0x41ea94 MADD	X28, X13, X7, XZR

0x41ea98 LDR	D0, [X26, X25,SXTW #3]    [6]

0x41ea9c ADD	X28, X12, X28,LSL #3

0x41eaa0 LDR	D2, [X28, X6,SXTW #3]    [8]

0x41eaa4 MADD	X6, X15, X20, XZR

0x41eaa8 ADD	X6, X14, X6,LSL #3

0x41eaac MADD	W5, W3, W5, W21

0x41eab0 ADD	W21, W5, #2

0x41eab4 SBFM	X21, X21, #61, #31

0x41eab8 ADD	X30, X30, X21

0x41eabc ADD	W5, W5, #1

0x41eac0 ADD	X27, X27, X21

0x41eac4 ADD	X6, X6, X21

0x41eac8 LD1	{V1.D[1]}, [X30]    [4]

0x41eacc MADD	X30, X13, X20, XZR

0x41ead0 LD1	{V0.D[1]}, [X27]    [7]

0x41ead4 ADD	X30, X12, X30,LSL #3

0x41ead8 ADD	X5, X30, W5,SXTW #3

0x41eadc LD1	{V2.D[1]}, [X5]    [2]

0x41eae0 ADD	X5, X30, X21

0x41eae4 FMLA	V2.2D, V1.2D, V0.2D

0x41eae8 LDR	D0, [X28, X25,SXTW #3]    [8]

0x41eaec LD1	{V0.D[1]}, [X5]    [1]

0x41eaf0 MADD	X5, X15, X7, XZR

0x41eaf4 ADD	X5, X14, X5,LSL #3

0x41eaf8 LDR	D1, [X5, X25,SXTW #3]    [5]

0x41eafc FSUB	V0.2D, V2.2D, V0.2D

0x41eb00 LD1	{V1.D[1]}, [X6]    [3]

0x41eb04 FDIV	V0.2D, V0.2D, V1.2D

0x41eb08 STR	D0, [X26, X25,SXTW #3]    [6]

0x41eb0c ST1	{V0.D[1]}, [X27]    [7]

0x41eb10 CMP	X2, X1

0x41eb14 B.NE	41ea40

/home/hbollore/qaas/qaas-runs/174-169-1878/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_mom.cpp: 147 - 149

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

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

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

149:         vel1(i, j) = (vel1(i, j) * node_mass_pre(i, j) + mom_flux(i - 1, j + 0) - mom_flux(i, j)) / node_mass_post(i, j);

/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	3.37 - 6.72
CQA speedup if FP arith vectorized	1.00
CQA speedup if fully vectorized	1.00
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.12 - 3.92
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.14]
Source	advec_mom.cpp:147-149,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	13.50 - 46.99
CQA cycles if no scalar integer	4.00 - 6.99
CQA cycles if FP arith vectorized	13.50 - 46.99
CQA cycles if fully vectorized	13.50 - 46.99
Front-end cycles	6.75
P0 cycles	0.50
P1 cycles	0.50
P2 cycles	5.33
P3 cycles	5.17
P4 cycles	5.17
P5 cycles	5.17
P6 cycles	12.00
P7 cycles	5.17
P8 cycles	2.50
P9 cycles	2.50
P10 cycles	2.50
P11 cycles	2.50
P12 cycles	4.00
P13 cycles	4.00
P14 cycles	4.00
P15 cycles	0.00
P16 cycles	0.00
DIV/SQRT cycles	13.50 - 46.99
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	0.59 - 0.17
Nb FLOP add-sub	2.00
Nb FLOP mul	0.00
Nb FLOP fma	2.00
Nb FLOP div	2.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	2.21 - 7.71
Bytes prefetched	0.00
Bytes loaded	80.00
Bytes stored	24.00
Stride 0	0.00
Stride 1	0.00
Stride n	0.00
Stride unknown	5.00
Stride indirect	4.00
Vectorization ratio all	13.64
Vectorization ratio load	0.00
Vectorization ratio store	0.00
Vectorization ratio mul	NA
Vectorization ratio add_sub	20.00
Vectorization ratio fma	100.00
Vectorization ratio div_sqrt	100.00
Vectorization ratio other	0.00
Vector-efficiency ratio all	59.09
Vector-efficiency ratio load	50.00
Vector-efficiency ratio store	50.00
Vector-efficiency ratio mul	NA
Vector-efficiency ratio add_sub	40.00
Vector-efficiency ratio fma	100.00
Vector-efficiency ratio div_sqrt	100.00
Vector-efficiency ratio other	100.00

Metric	Value
CQA speedup if no scalar integer	3.37 - 6.72
CQA speedup if FP arith vectorized	1.00
CQA speedup if fully vectorized	1.00
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.12 - 3.92
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.14]
Source	advec_mom.cpp:147-149,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	13.50 - 46.99
CQA cycles if no scalar integer	4.00 - 6.99
CQA cycles if FP arith vectorized	13.50 - 46.99
CQA cycles if fully vectorized	13.50 - 46.99
Front-end cycles	6.75
P0 cycles	0.50
P1 cycles	0.50
P2 cycles	5.33
P3 cycles	5.17
P4 cycles	5.17
P5 cycles	5.17
P6 cycles	12.00
P7 cycles	5.17
P8 cycles	2.50
P9 cycles	2.50
P10 cycles	2.50
P11 cycles	2.50
P12 cycles	4.00
P13 cycles	4.00
P14 cycles	4.00
P15 cycles	0.00
P16 cycles	0.00
DIV/SQRT cycles	13.50 - 46.99
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	0.59 - 0.17
Nb FLOP add-sub	2.00
Nb FLOP mul	0.00
Nb FLOP fma	2.00
Nb FLOP div	2.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	2.21 - 7.71
Bytes prefetched	0.00
Bytes loaded	80.00
Bytes stored	24.00
Stride 0	0.00
Stride 1	0.00
Stride n	0.00
Stride unknown	5.00
Stride indirect	4.00
Vectorization ratio all	13.64
Vectorization ratio load	0.00
Vectorization ratio store	0.00
Vectorization ratio mul	NA
Vectorization ratio add_sub	20.00
Vectorization ratio fma	100.00
Vectorization ratio div_sqrt	100.00
Vectorization ratio other	0.00
Vector-efficiency ratio all	59.09
Vector-efficiency ratio load	50.00
Vector-efficiency ratio store	50.00
Vector-efficiency ratio mul	NA
Vector-efficiency ratio add_sub	40.00
Vector-efficiency ratio fma	100.00
Vector-efficiency ratio div_sqrt	100.00
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.14]
Source file and lines	advec_mom.cpp:147-149
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:147-149
/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 46.99 to 6.99 cycles (6.72x 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 59% of vector register length is used (average across all VPU instructions).

Details

13% 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.
20% of VPU addition or subtraction 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

Detected 2 FMA (fused multiply-add) operations.

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

The binary loop is composed of 8 FP arithmetical operations:

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

The binary loop is loading 80 bytes. The binary loop is storing 24 bytes.

Arithmetic intensity

Arithmetic intensity is 0.08 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	5.33	5.17	5.17	5.17	12.00	5.17	2.50	2.50	2.50	2.50	4.00	4.00	4.00	0.00	0.00
cycles	0.50	0.50	5.33	5.17	5.17	5.17	12.00	5.17	2.50	2.50	2.50	2.50	4.00	4.00	4.00	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	13.50-46.99
Longest recurrence chain latency (RecMII)	1.00

Cycles summary

Front-end	6.75
Data deps.	1.00
Overall L1	13.50-46.99

Vectorization ratios

INT

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

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

INT+FP

all	13%
load	0%
store	0%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	20%
fma	100%
div/sqrt	100%
other	0%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 46.99 cycles. At this rate:

1% of peak load performance is reached (1.70 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
1% of peak store performance is reached (0.51 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the loop is: 41ea40

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput	Vectorization
ADD X5, X18, X1	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 W21, W4, W1	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 X1, X1, #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
SDIV X6, X5, X24	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	5-20	5-20	N/A
ADD X5, X5, #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
ADD W7, W22, W6	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 X5, X5, X24	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	5-20	5-20	N/A
SBFM X7, X7, #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 W20, W22, 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%)
MADD X28, X11, X7, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
SBFM X20, X20, #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 X26, X9, X7, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X30, X11, X20, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X27, X9, X20, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X28, X10, 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 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
ADD X30, X10, 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	N/A
ADD X27, X8, X27,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 W6, W3, W6, W21	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD W25, W6, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
LDR D1, [X28, X25,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 X28, X13, X7, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
LDR D0, [X26, X25,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 X28, X12, 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
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%)
MADD X6, X15, X20, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X6, X14, X6,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 W5, W3, W5, W21	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD W21, W5, #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	scal (25.0%)
SBFM X21, X21, #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 X30, X30, X21	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 W5, W5, #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
ADD X27, X27, X21	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, X6, X21	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]}, [X30]	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 X30, X13, X20, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
LD1 {V0.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%)
ADD X30, X12, 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	N/A
ADD X5, X30, W5,SXTW #3	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	2	0.50	N/A
LD1 {V2.D[1]}, [X5]	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 X5, X30, X21	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
FMLA V2.2D, V1.2D, V0.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	vect (100.0%)
LDR D0, [X28, X25,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 {V0.D[1]}, [X5]	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 X5, X15, X7, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X5, X14, X5,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, [X5, X25,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%)
FSUB V0.2D, V2.2D, 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%)
LD1 {V1.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%)
FDIV V0.2D, V0.2D, 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, [X26, X25,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]}, [X27]	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 X2, X1	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 41ea40 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii.omp_outlined.14+0x110>	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.14]
Source file and lines	advec_mom.cpp:147-149
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:147-149
/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 46.99 to 6.99 cycles (6.72x 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 59% of vector register length is used (average across all VPU instructions).

Details

13% 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.
20% of VPU addition or subtraction 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

Detected 2 FMA (fused multiply-add) operations.

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

The binary loop is composed of 8 FP arithmetical operations:

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

The binary loop is loading 80 bytes. The binary loop is storing 24 bytes.

Arithmetic intensity

Arithmetic intensity is 0.08 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	5.33	5.17	5.17	5.17	12.00	5.17	2.50	2.50	2.50	2.50	4.00	4.00	4.00	0.00	0.00
cycles	0.50	0.50	5.33	5.17	5.17	5.17	12.00	5.17	2.50	2.50	2.50	2.50	4.00	4.00	4.00	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	13.50-46.99
Longest recurrence chain latency (RecMII)	1.00

Cycles summary

Front-end	6.75
Data deps.	1.00
Overall L1	13.50-46.99

Vectorization ratios

INT

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

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

INT+FP

all	13%
load	0%
store	0%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	20%
fma	100%
div/sqrt	100%
other	0%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 46.99 cycles. At this rate:

1% of peak load performance is reached (1.70 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
1% of peak store performance is reached (0.51 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the loop is: 41ea40

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput	Vectorization
ADD X5, X18, X1	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 W21, W4, W1	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 X1, X1, #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
SDIV X6, X5, X24	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	5-20	5-20	N/A
ADD X5, X5, #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
ADD W7, W22, W6	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 X5, X5, X24	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	5-20	5-20	N/A
SBFM X7, X7, #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 W20, W22, 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%)
MADD X28, X11, X7, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
SBFM X20, X20, #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 X26, X9, X7, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X30, X11, X20, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X27, X9, X20, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X28, X10, 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 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
ADD X30, X10, 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	N/A
ADD X27, X8, X27,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 W6, W3, W6, W21	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD W25, W6, #1	1	0	0	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	1	0.17	N/A
LDR D1, [X28, X25,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 X28, X13, X7, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
LDR D0, [X26, X25,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 X28, X12, 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
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%)
MADD X6, X15, X20, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X6, X14, X6,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 W5, W3, W5, W21	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD W21, W5, #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	scal (25.0%)
SBFM X21, X21, #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 X30, X30, X21	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 W5, W5, #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
ADD X27, X27, X21	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, X6, X21	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]}, [X30]	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 X30, X13, X20, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
LD1 {V0.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%)
ADD X30, X12, 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	N/A
ADD X5, X30, W5,SXTW #3	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	2	0.50	N/A
LD1 {V2.D[1]}, [X5]	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 X5, X30, X21	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
FMLA V2.2D, V1.2D, V0.2D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	vect (100.0%)
LDR D0, [X28, X25,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 {V0.D[1]}, [X5]	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 X5, X15, X7, XZR	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X5, X14, X5,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, [X5, X25,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%)
FSUB V0.2D, V2.2D, 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%)
LD1 {V1.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%)
FDIV V0.2D, V0.2D, 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, [X26, X25,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]}, [X27]	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 X2, X1	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 41ea40 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii.omp_outlined.14+0x110>	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

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

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