OV - exec - Loop 196

	Loop Id: 196	Module: exec	Source: advec_mom.cpp:168-172 [...]	Coverage: 2.94%

0x41e7c0 MOVPRFX	Z26, Z25

0x41e7c4 SDIV	Z26.D, P0/M, Z26.D, Z1.D

0x41e7c8 DECD	X3, ALL

0x41e7cc ADD	Z27.D, Z2.D, Z26.D

0x41e7d0 MSB	Z26.D, P0/M, Z1.D, Z25.D

0x41e7d4 ADD	Z25.D, Z25.D, Z0.D

0x41e7d8 ADD	Z29.D, Z27.D, Z18.D

0x41e7dc SXTW	Z27.D, P0/M, Z27.D

0x41e7e0 MUL	Z10.D, Z4.D, Z27.D

0x41e7e4 MUL	Z12.D, Z5.D, Z27.D

0x41e7e8 SXTW	Z29.D, P0/M, Z29.D

0x41e7ec ADD	Z28.D, Z3.D, Z26.D

0x41e7f0 ADD	Z26.D, Z6.D, Z26.D

0x41e7f4 MUL	Z30.D, Z4.D, Z29.D

0x41e7f8 MUL	Z8.D, Z5.D, Z29.D

0x41e7fc MUL	Z29.D, Z16.D, Z29.D

0x41e800 ADR	Z10.D, [Z19, Z10.D,LSL #3]    [23]

0x41e804 ADR	Z12.D, [Z20, Z12.D,LSL #3]    [2]

0x41e808 ADR	Z11.D, [Z10, Z28.D,SXTW #3]    [1]

0x41e80c ADR	Z13.D, [Z12, Z28.D,SXTW #3]    [11]

0x41e810 ADR	Z30.D, [Z19, Z30.D,LSL #3]    [23]

0x41e814 ADR	Z8.D, [Z20, Z8.D,LSL #3]    [2]

0x41e818 ADR	Z29.D, [Z23, Z29.D,LSL #3]    [16]

0x41e81c LD1D	{Z11.D}, P0/Z, [V11.D]    [7]

0x41e820 LD1D	{Z13.D}, P0/Z, [V13.D]    [15]

0x41e824 ADR	Z31.D, [Z30, Z28.D,SXTW #3]    [6]

0x41e828 ADR	Z9.D, [Z8, Z28.D,SXTW #3]    [10]

0x41e82c ADR	Z30.D, [Z30, Z26.D,SXTW #3]    [6]

0x41e830 ADR	Z8.D, [Z8, Z26.D,SXTW #3]    [10]

0x41e834 ADR	Z29.D, [Z29, Z28.D,SXTW #3]    [25]

0x41e838 LD1D	{Z31.D}, P0/Z, [V31.D]    [8]

0x41e83c FMUL	Z11.D, Z11.D, Z13.D

0x41e840 LD1D	{Z9.D}, P0/Z, [V9.D]    [14]

0x41e844 LD1D	{Z30.D}, P0/Z, [V30.D]    [5]

0x41e848 LD1D	{Z8.D}, P0/Z, [V8.D]    [9]

0x41e84c FMAD	Z31.D, P0/M, Z9.D, Z11.D

0x41e850 ADR	Z9.D, [Z10, Z26.D,SXTW #3]    [1]

0x41e854 ADR	Z26.D, [Z12, Z26.D,SXTW #3]    [11]

0x41e858 LD1D	{Z9.D}, P0/Z, [V9.D]    [13]

0x41e85c LD1D	{Z26.D}, P0/Z, [V26.D]    [20]

0x41e860 FMAD	Z30.D, P0/M, Z8.D, Z31.D

0x41e864 FMAD	Z26.D, P0/M, Z9.D, Z30.D

0x41e868 MUL	Z30.D, Z7.D, Z27.D

0x41e86c FMUL	Z26.D, Z26.D, Z21.D

0x41e870 ADR	Z30.D, [Z22, Z30.D,LSL #3]    [12]

0x41e874 ADR	Z30.D, [Z30, Z28.D,SXTW #3]    [4]

0x41e878 ST1D	{Z26.D}, P0, [V30.D]    [3]

0x41e87c MUL	Z30.D, Z16.D, Z27.D

0x41e880 MUL	Z27.D, Z17.D, Z27.D

0x41e884 LD1D	{Z29.D}, P0/Z, [V29.D]    [24]

0x41e888 ADR	Z30.D, [Z23, Z30.D,LSL #3]    [16]

0x41e88c ADR	Z27.D, [Z24, Z27.D,LSL #3]    [17]

0x41e890 ADR	Z30.D, [Z30, Z28.D,SXTW #3]    [19]

0x41e894 ADR	Z27.D, [Z27, Z28.D,SXTW #3]    [22]

0x41e898 FSUB	Z26.D, Z26.D, Z29.D

0x41e89c LD1D	{Z29.D}, P0/Z, [V30.D]    [18]

0x41e8a0 FADD	Z26.D, Z26.D, Z29.D

0x41e8a4 ST1D	{Z26.D}, P0, [V27.D]    [21]

0x41e8a8 CBNZ	X3, 41e7c0

/home/hbollore/qaas/qaas-runs/174-169-1878/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_mom.cpp: 168 - 172

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

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

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

170:           node_mass_post(i, j) = 0.25 * (density1(i + 0, j - 1) * post_vol(i + 0, j - 1) + density1(i, j) * post_vol(i, j) +

171:                                          density1(i - 1, j - 1) * post_vol(i - 1, j - 1) + density1(i - 1, j + 0) * post_vol(i - 1, j + 0));

172:           node_mass_pre(i, j) = node_mass_post(i, j) - node_flux(i + 0, j - 1) + node_flux(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	1.00
CQA speedup if FP arith vectorized	1.00
CQA speedup if fully vectorized	1.26 - 1.00
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.92 - 1.63
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.18]
Source	advec_mom.cpp:168-172,context.h:69-69
Source loop unroll info	unrolled by 2
Source loop unroll confidence level	max
Unroll/vectorization loop type	NA
Unroll factor	2
CQA cycles	12.25 - 20.00
CQA cycles if no scalar integer	12.25 - 20.00
CQA cycles if FP arith vectorized	12.25 - 20.00
CQA cycles if fully vectorized	9.75 - 20.00
Front-end cycles	6.38
P0 cycles	0.50
P1 cycles	0.50
P2 cycles	0.00
P3 cycles	0.00
P4 cycles	0.00
P5 cycles	0.00
P6 cycles	0.50
P7 cycles	0.50
P8 cycles	12.25
P9 cycles	12.25
P10 cycles	12.25
P11 cycles	12.25
P12 cycles	4.00
P13 cycles	4.00
P14 cycles	4.00
P15 cycles	0.00
P16 cycles	0.00
DIV/SQRT cycles	0.87 - 20.00
Inter-iter dependencies cycles	2
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	59.00
Nb uops	51.00
Nb loads	NA
Nb stores	2.00
Nb stack references	0.00
FLOP/cycle	1.63 - 1.00
Nb FLOP add-sub	4.00
Nb FLOP mul	4.00
Nb FLOP fma	6.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	9.60 - 15.67
Bytes prefetched	0.00
Bytes loaded	160.00
Bytes stored	32.00
Stride 0	0.00
Stride 1	0.00
Stride n	0.00
Stride unknown	17.00
Stride indirect	8.00
Vectorization ratio all	62.07
Vectorization ratio load	33.33
Vectorization ratio store	100.00
Vectorization ratio mul	100.00
Vectorization ratio add_sub	25.00
Vectorization ratio fma	100.00
Vectorization ratio div_sqrt	100.00
Vectorization ratio other	66.67
Vector-efficiency ratio all	81.90
Vector-efficiency ratio load	66.67
Vector-efficiency ratio store	100.00
Vector-efficiency ratio mul	100.00
Vector-efficiency ratio add_sub	62.50
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	1.00
CQA speedup if FP arith vectorized	1.00
CQA speedup if fully vectorized	1.26 - 1.00
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.92 - 1.63
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.18]
Source	advec_mom.cpp:168-172,context.h:69-69
Source loop unroll info	unrolled by 2
Source loop unroll confidence level	max
Unroll/vectorization loop type	NA
Unroll factor	2
CQA cycles	12.25 - 20.00
CQA cycles if no scalar integer	12.25 - 20.00
CQA cycles if FP arith vectorized	12.25 - 20.00
CQA cycles if fully vectorized	9.75 - 20.00
Front-end cycles	6.38
P0 cycles	0.50
P1 cycles	0.50
P2 cycles	0.00
P3 cycles	0.00
P4 cycles	0.00
P5 cycles	0.00
P6 cycles	0.50
P7 cycles	0.50
P8 cycles	12.25
P9 cycles	12.25
P10 cycles	12.25
P11 cycles	12.25
P12 cycles	4.00
P13 cycles	4.00
P14 cycles	4.00
P15 cycles	0.00
P16 cycles	0.00
DIV/SQRT cycles	0.87 - 20.00
Inter-iter dependencies cycles	2
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	59.00
Nb uops	51.00
Nb loads	NA
Nb stores	2.00
Nb stack references	0.00
FLOP/cycle	1.63 - 1.00
Nb FLOP add-sub	4.00
Nb FLOP mul	4.00
Nb FLOP fma	6.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	9.60 - 15.67
Bytes prefetched	0.00
Bytes loaded	160.00
Bytes stored	32.00
Stride 0	0.00
Stride 1	0.00
Stride n	0.00
Stride unknown	17.00
Stride indirect	8.00
Vectorization ratio all	62.07
Vectorization ratio load	33.33
Vectorization ratio store	100.00
Vectorization ratio mul	100.00
Vectorization ratio add_sub	25.00
Vectorization ratio fma	100.00
Vectorization ratio div_sqrt	100.00
Vectorization ratio other	66.67
Vector-efficiency ratio all	81.90
Vector-efficiency ratio load	66.67
Vector-efficiency ratio store	100.00
Vector-efficiency ratio mul	100.00
Vector-efficiency ratio add_sub	62.50
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.18]
Source file and lines	advec_mom.cpp:168-172
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:168-172
/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 2 (including vectorization).

gain
potential
hint
expert

Vectorization

Your loop is partially vectorized. 81% of vector register length is used (average across all VPU instructions).

Details

62% of VPU instructions are used in vector version (process two or more data elements in vector registers):

33% of VPU loads are used in vector version.
25% of VPU addition or subtraction instructions are used in vector version.
66% of VPU instructions that are not load, store, addition, subtraction nor multiply instructions are used in vector version.

FMA

Detected 6 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:

10: addition or subtraction (6 inside FMA instructions)
10: multiply (6 inside FMA instructions)

The binary loop is loading 160 bytes. The binary loop is storing 32 bytes.

Arithmetic intensity

Arithmetic intensity is 0.10 FP operations per loaded or stored byte.

General properties

nb instructions	59
loop length	236
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

6.38 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	0.00	0.00	0.00	0.00	0.50	0.50	12.25	12.25	12.25	12.25	4.00	4.00	4.00	0.00	0.00
cycles	0.50	0.50	0.00	0.00	0.00	0.00	0.50	0.50	12.25	12.25	12.25	12.25	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	0.87-20.00
Longest recurrence chain latency (RecMII)	2.00

Cycles summary

Front-end	6.38
Data deps.	2.00
Overall L1	12.25-20.00

Vectorization ratios

INT

all	56%
load	33%
store	100%
mul	100%
add-sub	19%
fma	100%
other	75%

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	NA (no div/sqrt vectorizable/vectorized instructions)
other	NA (no other vectorizable/vectorized instructions)

INT+FP

all	62%
load	33%
store	100%
mul	100%
add-sub	25%
fma	100%
div/sqrt	100%
other	66%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 20.00 cycles. At this rate:

8% of peak load performance is reached (8.00 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
3% of peak store performance is reached (1.60 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the loop is: 41e7c0

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15	P16	Latency	Recip. throughput	Vectorization
MOVPRFX Z26, Z25	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (100.0%)
SDIV Z26.D, P0/M, Z26.D, Z1.D	1	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	7-20	0.87-20	vect (100.0%)
DECD X3, ALL	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	0	0	2	0.50	scal (50.0%)
ADD Z27.D, Z2.D, Z26.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	vect (100.0%)
MSB Z26.D, P0/M, Z1.D, Z25.D	1	0	0	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	0	0	5	1	vect (100.0%)
ADD Z25.D, Z25.D, Z0.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	vect (100.0%)
ADD Z29.D, Z27.D, Z18.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	vect (100.0%)
SXTW Z27.D, P0/M, Z27.D	1	0	0	0	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	0	2	0.50	vect (100.0%)
MUL Z10.D, Z4.D, Z27.D																					vect (100.0%)
MUL Z12.D, Z5.D, Z27.D																					vect (100.0%)
SXTW Z29.D, P0/M, Z29.D	1	0	0	0	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	0	2	0.50	vect (100.0%)
ADD Z28.D, Z3.D, Z26.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	vect (100.0%)
ADD Z26.D, Z6.D, Z26.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	vect (100.0%)
MUL Z30.D, Z4.D, Z29.D																					vect (100.0%)
MUL Z8.D, Z5.D, Z29.D																					vect (100.0%)
MUL Z29.D, Z16.D, Z29.D																					vect (100.0%)
ADR Z10.D, [Z19, Z10.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z12.D, [Z20, Z12.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z11.D, [Z10, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z13.D, [Z12, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z30.D, [Z19, Z30.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z8.D, [Z20, Z8.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z29.D, [Z23, Z29.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
LD1D {Z11.D}, P0/Z, [V11.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
LD1D {Z13.D}, P0/Z, [V13.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
ADR Z31.D, [Z30, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z9.D, [Z8, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z30.D, [Z30, Z26.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z8.D, [Z8, Z26.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z29.D, [Z29, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
LD1D {Z31.D}, P0/Z, [V31.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
FMUL Z11.D, Z11.D, Z13.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	3	0.25	vect (100.0%)
LD1D {Z9.D}, P0/Z, [V9.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
LD1D {Z30.D}, P0/Z, [V30.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
LD1D {Z8.D}, P0/Z, [V8.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
FMAD Z31.D, P0/M, Z9.D, Z11.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	4	0.25	vect (100.0%)
ADR Z9.D, [Z10, Z26.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z26.D, [Z12, Z26.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
LD1D {Z9.D}, P0/Z, [V9.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
LD1D {Z26.D}, P0/Z, [V26.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
FMAD Z30.D, P0/M, Z8.D, Z31.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	4	0.25	vect (100.0%)
FMAD Z26.D, P0/M, Z9.D, Z30.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	4	0.25	vect (100.0%)
MUL Z30.D, Z7.D, Z27.D																					vect (100.0%)
FMUL Z26.D, Z26.D, Z21.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	3	0.25	vect (100.0%)
ADR Z30.D, [Z22, Z30.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z30.D, [Z30, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ST1D {Z26.D}, P0, [V30.D]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	1	vect (100.0%)
MUL Z30.D, Z16.D, Z27.D																					vect (100.0%)
MUL Z27.D, Z17.D, Z27.D																					vect (100.0%)
LD1D {Z29.D}, P0/Z, [V29.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
ADR Z30.D, [Z23, Z30.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z27.D, [Z24, Z27.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z30.D, [Z30, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z27.D, [Z27, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
FSUB Z26.D, Z26.D, Z29.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	vect (100.0%)
LD1D {Z29.D}, P0/Z, [V30.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
FADD Z26.D, Z26.D, Z29.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	vect (100.0%)
ST1D {Z26.D}, P0, [V27.D]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	1	vect (100.0%)
CBNZ X3, 41e7c0 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii.omp_outlined.18+0x160>	1	0.50	0.50	0	0	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.18]
Source file and lines	advec_mom.cpp:168-172
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:168-172
/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 2 (including vectorization).

gain
potential
hint
expert

Vectorization

Your loop is partially vectorized. 81% of vector register length is used (average across all VPU instructions).

Details

62% of VPU instructions are used in vector version (process two or more data elements in vector registers):

33% of VPU loads are used in vector version.
25% of VPU addition or subtraction instructions are used in vector version.
66% of VPU instructions that are not load, store, addition, subtraction nor multiply instructions are used in vector version.

FMA

Detected 6 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:

10: addition or subtraction (6 inside FMA instructions)
10: multiply (6 inside FMA instructions)

The binary loop is loading 160 bytes. The binary loop is storing 32 bytes.

Arithmetic intensity

Arithmetic intensity is 0.10 FP operations per loaded or stored byte.

General properties

nb instructions	59
loop length	236
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

6.38 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	0.00	0.00	0.00	0.00	0.50	0.50	12.25	12.25	12.25	12.25	4.00	4.00	4.00	0.00	0.00
cycles	0.50	0.50	0.00	0.00	0.00	0.00	0.50	0.50	12.25	12.25	12.25	12.25	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	0.87-20.00
Longest recurrence chain latency (RecMII)	2.00

Cycles summary

Front-end	6.38
Data deps.	2.00
Overall L1	12.25-20.00

Vectorization ratios

INT

all	56%
load	33%
store	100%
mul	100%
add-sub	19%
fma	100%
other	75%

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	NA (no div/sqrt vectorizable/vectorized instructions)
other	NA (no other vectorizable/vectorized instructions)

INT+FP

all	62%
load	33%
store	100%
mul	100%
add-sub	25%
fma	100%
div/sqrt	100%
other	66%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 20.00 cycles. At this rate:

8% of peak load performance is reached (8.00 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
3% of peak store performance is reached (1.60 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the loop is: 41e7c0

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15	P16	Latency	Recip. throughput	Vectorization
MOVPRFX Z26, Z25	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (100.0%)
SDIV Z26.D, P0/M, Z26.D, Z1.D	1	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	7-20	0.87-20	vect (100.0%)
DECD X3, ALL	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	0	0	2	0.50	scal (50.0%)
ADD Z27.D, Z2.D, Z26.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	vect (100.0%)
MSB Z26.D, P0/M, Z1.D, Z25.D	1	0	0	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	0	0	5	1	vect (100.0%)
ADD Z25.D, Z25.D, Z0.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	vect (100.0%)
ADD Z29.D, Z27.D, Z18.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	vect (100.0%)
SXTW Z27.D, P0/M, Z27.D	1	0	0	0	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	0	2	0.50	vect (100.0%)
MUL Z10.D, Z4.D, Z27.D																					vect (100.0%)
MUL Z12.D, Z5.D, Z27.D																					vect (100.0%)
SXTW Z29.D, P0/M, Z29.D	1	0	0	0	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	0	2	0.50	vect (100.0%)
ADD Z28.D, Z3.D, Z26.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	vect (100.0%)
ADD Z26.D, Z6.D, Z26.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	vect (100.0%)
MUL Z30.D, Z4.D, Z29.D																					vect (100.0%)
MUL Z8.D, Z5.D, Z29.D																					vect (100.0%)
MUL Z29.D, Z16.D, Z29.D																					vect (100.0%)
ADR Z10.D, [Z19, Z10.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z12.D, [Z20, Z12.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z11.D, [Z10, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z13.D, [Z12, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z30.D, [Z19, Z30.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z8.D, [Z20, Z8.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z29.D, [Z23, Z29.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
LD1D {Z11.D}, P0/Z, [V11.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
LD1D {Z13.D}, P0/Z, [V13.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
ADR Z31.D, [Z30, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z9.D, [Z8, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z30.D, [Z30, Z26.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z8.D, [Z8, Z26.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z29.D, [Z29, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
LD1D {Z31.D}, P0/Z, [V31.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
FMUL Z11.D, Z11.D, Z13.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	3	0.25	vect (100.0%)
LD1D {Z9.D}, P0/Z, [V9.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
LD1D {Z30.D}, P0/Z, [V30.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
LD1D {Z8.D}, P0/Z, [V8.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
FMAD Z31.D, P0/M, Z9.D, Z11.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	4	0.25	vect (100.0%)
ADR Z9.D, [Z10, Z26.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z26.D, [Z12, Z26.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
LD1D {Z9.D}, P0/Z, [V9.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
LD1D {Z26.D}, P0/Z, [V26.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
FMAD Z30.D, P0/M, Z8.D, Z31.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	4	0.25	vect (100.0%)
FMAD Z26.D, P0/M, Z9.D, Z30.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	4	0.25	vect (100.0%)
MUL Z30.D, Z7.D, Z27.D																					vect (100.0%)
FMUL Z26.D, Z26.D, Z21.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	3	0.25	vect (100.0%)
ADR Z30.D, [Z22, Z30.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z30.D, [Z30, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ST1D {Z26.D}, P0, [V30.D]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	1	vect (100.0%)
MUL Z30.D, Z16.D, Z27.D																					vect (100.0%)
MUL Z27.D, Z17.D, Z27.D																					vect (100.0%)
LD1D {Z29.D}, P0/Z, [V29.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
ADR Z30.D, [Z23, Z30.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z27.D, [Z24, Z27.D,LSL #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z30.D, [Z30, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
ADR Z27.D, [Z27, Z28.D,SXTW #3]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (50.0%)
FSUB Z26.D, Z26.D, Z29.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	vect (100.0%)
LD1D {Z29.D}, P0/Z, [V30.D]	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.33	0.33	0.33	0	0	9	1	vect (100.0%)
FADD Z26.D, Z26.D, Z29.D	1	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	vect (100.0%)
ST1D {Z26.D}, P0, [V27.D]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	1	vect (100.0%)
CBNZ X3, 41e7c0 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii.omp_outlined.18+0x160>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A

Report Configuration

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

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