OV - exec - Loop 378

0x4194e0 LDR	X8, [X19, #512]    [4]

0x4194e4 LDR	X5, [X19, #464]    [4]

0x4194e8 ORR	W1, WZR, W20

0x4194ec ORR	X4, XZR, X22

0x4194f0 ADD	X8, X8, X25

0x4194f4 LDR	X0, [X8, X24,LSL #3]    [1]

0x4194f8 LDR	X8, [X21, #2712]    [3]

0x4194fc ADD	X8, X8, X24,LSL #2

0x419500 ADD	X24, X24, #1

0x419504 LDP	W2, W3, [X8]    [2]

0x419508 BL	416f40

0x41950c LDR	X8, [X19, #160]    [4]

0x419510 FADD	D8, D0, D8

0x419514 CMP	X8, X24

0x419518 B.HI	4194e0

/usr/lib/gcc/aarch64-linux-gnu/11/../../../../include/c++/11/bits/stl_vector.h: 1046 - 1169

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

1046: 	return *(this->_M_impl._M_start + __n);

[...]

1169:       { return _M_data_ptr(this->_M_impl._M_start); }

/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsVector.h: 229 - 229

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

229:     return X[i];

/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/QMCWaveFunctions/Jastrow/TwoBodyJastrowRef.h: 125 - 130

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

125:     for (int jg = 0; jg < NumGroups; ++jg)

126:     {

127:       const FuncType& f2(*F[igt + jg]);

128:       int iStart = P.first(jg);

129:       int iEnd   = P.last(jg);

130:       curUat += f2.evaluateV(iat, iStart, iEnd, dist, DistCompressed.data());

/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Particle/ParticleSet.h: 302 - 305

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

302:   inline int first(int igroup) const { return SubPtcl[igroup]; }

303:

304:   /// return the last index of a group i

305:   inline int last(int igroup) const { return SubPtcl[igroup + 1]; }

Coverage (%)	Name	Source Location	Module
►100.00+	qmcplusplus::WaveFunction::rat[...]	NewTimer.h:249	exec
○	.omp_outlined..62	miniqmc.cpp:485	exec
○	__kmp_invoke_microtask		libomp.so

Path /

Metric	Value
CQA speedup if no scalar integer	1.00
CQA speedup if FP arith vectorized	1.00
CQA speedup if fully vectorized	7.50
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.07
Bottlenecks
Function	miniqmcreference::TwoBodyJastrowRef >::ratio(qmcplusplus::ParticleSet&, int)
Source	stl_vector.h:1046-1046,stl_vector.h:1169-1169,OhmmsVector.h:229-229,TwoBodyJastrowRef.h:125-130,ParticleSet.h:302-305
Source loop unroll info	not unrolled or unrolled with no peel/tail loop
Source loop unroll confidence level	max
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	2.00
CQA cycles if no scalar integer	2.00
CQA cycles if FP arith vectorized	2.00
CQA cycles if fully vectorized	0.27
Front-end cycles	1.88
DIV/SQRT cycles	1.00
P0 cycles	1.00
P1 cycles	1.75
P2 cycles	1.75
P3 cycles	1.75
P4 cycles	1.75
P5 cycles	0.25
P6 cycles	0.25
P7 cycles	0.25
P8 cycles	0.25
P9 cycles	2.00
P10 cycles	2.00
P11 cycles	2.00
P12 cycles	0.00
P13 cycles	0.00
P14 cycles	0.00
Inter-iter dependencies cycles	2
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	15.00
Nb uops	15.00
Nb loads	NA
Nb stores	0.00
Nb stack references	0.00
FLOP/cycle	0.50
Nb FLOP add-sub	1.00
Nb FLOP mul	0.00
Nb FLOP fma	0.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	24.00
Bytes prefetched	0.00
Bytes loaded	48.00
Bytes stored	0.00
Stride 0	NA
Stride 1	NA
Stride n	NA
Stride unknown	NA
Stride indirect	NA
Vectorization ratio all	0.00
Vectorization ratio load	NA
Vectorization ratio store	NA
Vectorization ratio mul	NA
Vectorization ratio add_sub	NA
Vectorization ratio fma	NA
Vectorization ratio div_sqrt	NA
Vectorization ratio other	0.00
Vector-efficiency ratio all	18.75
Vector-efficiency ratio load	NA
Vector-efficiency ratio store	NA
Vector-efficiency ratio mul	NA
Vector-efficiency ratio add_sub	NA
Vector-efficiency ratio fma	NA
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	18.75

Metric	Value
CQA speedup if no scalar integer	1.00
CQA speedup if FP arith vectorized	1.00
CQA speedup if fully vectorized	7.50
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.07
Bottlenecks
Function	miniqmcreference::TwoBodyJastrowRef >::ratio(qmcplusplus::ParticleSet&, int)
Source	stl_vector.h:1046-1046,stl_vector.h:1169-1169,OhmmsVector.h:229-229,TwoBodyJastrowRef.h:125-130,ParticleSet.h:302-305
Source loop unroll info	not unrolled or unrolled with no peel/tail loop
Source loop unroll confidence level	max
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	2.00
CQA cycles if no scalar integer	2.00
CQA cycles if FP arith vectorized	2.00
CQA cycles if fully vectorized	0.27
Front-end cycles	1.88
DIV/SQRT cycles	1.00
P0 cycles	1.00
P1 cycles	1.75
P2 cycles	1.75
P3 cycles	1.75
P4 cycles	1.75
P5 cycles	0.25
P6 cycles	0.25
P7 cycles	0.25
P8 cycles	0.25
P9 cycles	2.00
P10 cycles	2.00
P11 cycles	2.00
P12 cycles	0.00
P13 cycles	0.00
P14 cycles	0.00
Inter-iter dependencies cycles	2
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	15.00
Nb uops	15.00
Nb loads	NA
Nb stores	0.00
Nb stack references	0.00
FLOP/cycle	0.50
Nb FLOP add-sub	1.00
Nb FLOP mul	0.00
Nb FLOP fma	0.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	24.00
Bytes prefetched	0.00
Bytes loaded	48.00
Bytes stored	0.00
Stride 0	NA
Stride 1	NA
Stride n	NA
Stride unknown	NA
Stride indirect	NA
Vectorization ratio all	0.00
Vectorization ratio load	NA
Vectorization ratio store	NA
Vectorization ratio mul	NA
Vectorization ratio add_sub	NA
Vectorization ratio fma	NA
Vectorization ratio div_sqrt	NA
Vectorization ratio other	0.00
Vector-efficiency ratio all	18.75
Vector-efficiency ratio load	NA
Vector-efficiency ratio store	NA
Vector-efficiency ratio mul	NA
Vector-efficiency ratio add_sub	NA
Vector-efficiency ratio fma	NA
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	18.75

Path /

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

Function	miniqmcreference::TwoBodyJastrowRef >::ratio(qmcplusplus::ParticleSet&, int)
Source file and lines	TwoBodyJastrowRef.h:125-130
Module	exec

The loop is defined in:

/usr/lib/gcc/aarch64-linux-gnu/11/../../../../include/c++/11/bits/stl_vector.h:1046,1169
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsVector.h:229
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/QMCWaveFunctions/Jastrow/TwoBodyJastrowRef.h:125-130
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Particle/ParticleSet.h:302-305

The related source loop is not unrolled or unrolled with no peel/tail loop.

gain
potential
hint
expert

Vectorization

Your loop is not vectorized. Only 18% of vector register length is used (average across all VPU instructions). By vectorizing your loop, you can lower the cost of an iteration from 2.00 to 0.27 cycles (7.50x speedup).

Details

All VPU instructions are used in scalar version (process only one data element in vector registers). Since your execution units are vector units, only a vectorized loop can use their full power.

Workaround

Try another compiler or update/tune your current one
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)

No data for this section

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

The binary loop is composed of 1 FP arithmetical operations:

1: addition or subtraction

The binary loop is loading 48 bytes.

Arithmetic intensity

Arithmetic intensity is 0.02 FP operations per loaded or stored byte.

General properties

nb instructions	15
loop length	60
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

1.88 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	1.00	1.00	1.75	1.75	1.75	1.75	0.25	0.25	0.25	0.25	2.00	2.00	2.00	0.00	0.00
cycles	1.00	1.00	1.75	1.75	1.75	1.75	0.25	0.25	0.25	0.25	2.00	2.00	2.00	0.00	0.00

Cycles executing div or sqrt instructions	NA
Longest recurrence chain latency (RecMII)	2.00

Cycles summary

Front-end	1.88
Data deps.	2.00
Overall L1	2.00

Vectorization ratios

all	0%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	NA (no add-sub vectorizable/vectorized instructions)
fma	NA (no fma vectorizable/vectorized instructions)
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 2.00 cycles. At this rate:

25% of peak load performance is reached (24.00 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the loop is: 4194e0

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	Latency	Recip. throughput
LDR X8, [X19, #512]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
LDR X5, [X19, #464]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
ORR W1, WZR, W20	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ORR X4, XZR, X22	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD X8, X8, X25	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR X0, [X8, X24,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
LDR X8, [X21, #2712]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
ADD X8, X8, X24,LSL #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD X24, X24, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDP W2, W3, [X8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
BL 416f40 <_ZNK11qmcplusplus14BsplineFunctorIdE9evaluateVEiiiPKdPd>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	1	0.50
LDR X8, [X19, #160]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
FADD D8, D0, D8	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
CMP X8, X24	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.HI 4194e0 <_ZN16miniqmcreference17TwoBodyJastrowRefIN11qmcplusplus14BsplineFunctorIdEEE5ratioERNS1_11ParticleSetEi+0x60>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50

Function	miniqmcreference::TwoBodyJastrowRef >::ratio(qmcplusplus::ParticleSet&, int)
Source file and lines	TwoBodyJastrowRef.h:125-130
Module	exec

The loop is defined in:

/usr/lib/gcc/aarch64-linux-gnu/11/../../../../include/c++/11/bits/stl_vector.h:1046,1169
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsVector.h:229
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/QMCWaveFunctions/Jastrow/TwoBodyJastrowRef.h:125-130
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Particle/ParticleSet.h:302-305

The related source loop is not unrolled or unrolled with no peel/tail loop.

gain
potential
hint
expert

Vectorization

Your loop is not vectorized. Only 18% of vector register length is used (average across all VPU instructions). By vectorizing your loop, you can lower the cost of an iteration from 2.00 to 0.27 cycles (7.50x speedup).

Details

All VPU instructions are used in scalar version (process only one data element in vector registers). Since your execution units are vector units, only a vectorized loop can use their full power.

Workaround

Try another compiler or update/tune your current one
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)

No data for this section

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

The binary loop is composed of 1 FP arithmetical operations:

1: addition or subtraction

The binary loop is loading 48 bytes.

Arithmetic intensity

Arithmetic intensity is 0.02 FP operations per loaded or stored byte.

General properties

nb instructions	15
loop length	60
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

1.88 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	1.00	1.00	1.75	1.75	1.75	1.75	0.25	0.25	0.25	0.25	2.00	2.00	2.00	0.00	0.00
cycles	1.00	1.00	1.75	1.75	1.75	1.75	0.25	0.25	0.25	0.25	2.00	2.00	2.00	0.00	0.00

Cycles executing div or sqrt instructions	NA
Longest recurrence chain latency (RecMII)	2.00

Cycles summary

Front-end	1.88
Data deps.	2.00
Overall L1	2.00

Vectorization ratios

all	0%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	NA (no add-sub vectorizable/vectorized instructions)
fma	NA (no fma vectorizable/vectorized instructions)
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 2.00 cycles. At this rate:

25% of peak load performance is reached (24.00 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the loop is: 4194e0

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	Latency	Recip. throughput
LDR X8, [X19, #512]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
LDR X5, [X19, #464]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
ORR W1, WZR, W20	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ORR X4, XZR, X22	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD X8, X8, X25	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR X0, [X8, X24,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
LDR X8, [X21, #2712]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
ADD X8, X8, X24,LSL #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD X24, X24, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDP W2, W3, [X8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
BL 416f40 <_ZNK11qmcplusplus14BsplineFunctorIdE9evaluateVEiiiPKdPd>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	1	0.50
LDR X8, [X19, #160]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
FADD D8, D0, D8	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
CMP X8, X24	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.HI 4194e0 <_ZN16miniqmcreference17TwoBodyJastrowRefIN11qmcplusplus14BsplineFunctorIdEEE5ratioERNS1_11ParticleSetEi+0x60>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50

Report Configuration

Vectorization

Details

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

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