OV - exec - Loop 1375

	Loop Id: 1375	Module: exec	Source: OhmmsVector.h:178-248 [...]	Coverage: 0.01%

0x467198 ORR	W16, WZR, W3

0x46719c LDR	X17, [X19, #280]

0x4671a0 LDR	X30, [X19, #320]

0x4671a4 MADD	X21, X21, X17, XZR

0x4671a8 CMP	W3, #0

0x4671ac B.LE	4673bc

0x4671b0 SUB	W1, W3, #1

0x4671b4 ORR	W5, WZR, W3

0x4671b8 CMP	W1, #2

0x4671bc B.LS	4673c8

0x4671c0 ADD	X9, X21, X21,LSL #1

0x4671c4 UBFM	W12, W3, #2, #31

0x4671c8 MOVZ	W2, #96

0x4671cc MOVI	V26.2D, #0

0x4671d0 ORR	X10, XZR, X0

0x4671d4 ADD	X14, X30, X9,LSL #3

0x4671d8 UMADDL	X11, W12, W2, X14

0x4671dc ORR	V4.16B, V26.16B, V26.16B

0x4671e0 ORR	V22.16B, V26.16B, V26.16B

0x4671e4 ORR	V3.16B, V26.16B, V26.16B

0x4671e8 ORR	V19.16B, V26.16B, V26.16B

0x4671ec ORR	V5.16B, V26.16B, V26.16B

0x4671f0 SUB	X6, X11, X14

0x4671f4 TBZ	W6, #5, 467230

0x4671f8 LDP	Q28, Q29, [X0]

0x4671fc ADD	X10, X0, #32

0x467200 LDP	Q2, Q25, [X14, #16]

0x467204 LDP	Q24, Q30, [X14, #48]

0x467208 LDR	Q27, [X14, #80]

0x46720c LDR	Q6, [X14], #96

0x467210 FMLA	V19.2D, V28.2D, V2.2D

0x467214 FMLA	V3.2D, V25.2D, V28.D[1]

0x467218 FMLA	V22.2D, V24.2D, V29.D[0]

0x46721c FMLA	V4.2D, V29.2D, V30.2D

0x467220 FMLA	V5.2D, V6.2D, V28.D[0]

0x467224 FMLA	V26.2D, V27.2D, V29.D[1]

0x467228 CMP	X14, X11

0x46722c B.EQ	467298

  (1376) 0x467230 ORR	X13, XZR, X14

  (1376) 0x467234 LDP	Q16, Q17, [X10]

  (1376) 0x467238 LDR	Q7, [X13], #96

  (1376) 0x46723c LDP	Q18, Q23, [X14, #16]

  (1376) 0x467240 LDP	Q20, Q21, [X14, #48]

  (1376) 0x467244 LDP	Q0, Q1, [X14, #80]

  (1376) 0x467248 ADD	X14, X14, #192

  (1376) 0x46724c FMLA	V5.2D, V7.2D, V16.D[0]

  (1376) 0x467250 LDP	Q28, Q29, [X10, #32]

  (1376) 0x467254 ADD	X10, X10, #64

  (1376) 0x467258 LDP	Q31, Q15, [X13, #16]

  (1376) 0x46725c FMLA	V19.2D, V16.2D, V18.2D

  (1376) 0x467260 LDP	Q2, Q25, [X13, #48]

  (1376) 0x467264 LDR	Q24, [X13, #80]

  (1376) 0x467268 FMLA	V3.2D, V23.2D, V16.D[1]

  (1376) 0x46726c FMLA	V22.2D, V20.2D, V17.D[0]

  (1376) 0x467270 FMLA	V4.2D, V17.2D, V21.2D

  (1376) 0x467274 FMLA	V26.2D, V0.2D, V17.D[1]

  (1376) 0x467278 FMLA	V19.2D, V28.2D, V31.2D

  (1376) 0x46727c FMLA	V3.2D, V15.2D, V28.D[1]

  (1376) 0x467280 FMLA	V22.2D, V2.2D, V29.D[0]

  (1376) 0x467284 FMLA	V4.2D, V29.2D, V25.2D

  (1376) 0x467288 FMLA	V5.2D, V1.2D, V28.D[0]

  (1376) 0x46728c FMLA	V26.2D, V24.2D, V29.D[1]

  (1376) 0x467290 CMP	X14, X11

  (1376) 0x467294 B.NE	467230

0x467298 MOV	D27, V3.D[1]

0x46729c AND	W7, W5, #1886

0x4672a0 EXT	V3.16B, V19.16B, V3.16B, #8

0x4672a4 MOV	D6, V26.D[1]

0x4672a8 EXT	V26.16B, V4.16B, V26.16B, #8

0x4672ac FADD	V30.2D, V5.2D, V22.2D

0x4672b0 FADD	D19, D4, D19

0x4672b4 FADD	V22.2D, V3.2D, V26.2D

0x4672b8 FADD	D4, D27, D6

0x4672bc FADD	V31.2D, V22.2D, V30.2D

0x4672c0 FADD	D15, D4, D19

0x4672c4 ANDS	XZR, X5, #4160

0x4672c8 B.EQ	46735c

0x4672cc ADD	X15, X21, W7,SXTW

0x4672d0 ADD	X18, X15, X15,LSL #1

0x4672d4 ORR	W25, WZR, W7

0x4672d8 UBFM	X20, X25, #61, #60

0x4672dc ADD	W4, W7, #1

0x4672e0 UBFM	X8, X18, #61, #60

0x4672e4 ADD	X22, X30, X8

0x4672e8 LDR	D5, [X0, X20]

0x4672ec LDR	Q7, [X30, X8]

0x4672f0 LDR	D16, [X22, #16]

0x4672f4 FMLA	V31.2D, V7.2D, V5.D[0]

0x4672f8 FMADD	D15, D5, D16, D15

0x4672fc CMP	W16, W4

0x467300 B.LE	46735c

0x467304 ADD	X26, X21, W4,SXTW

0x467308 ADD	X3, X26, X26,LSL #1

0x46730c ADD	X17, X0, X20

0x467310 ADD	W1, W7, #2

0x467314 UBFM	X0, X3, #61, #60

0x467318 LDR	D23, [X17, #8]

0x46731c ADD	X5, X30, X0

0x467320 LDR	Q17, [X30, X0]

0x467324 LDR	D18, [X5, #16]

0x467328 FMLA	V31.2D, V17.2D, V23.D[0]

0x46732c FMADD	D15, D23, D18, D15

0x467330 CMP	W16, W1

0x467334 B.LE	46735c

0x467338 ADD	X16, X21, W1,SXTW

0x46733c ADD	X21, X16, X16,LSL #1

0x467340 LDR	D0, [X17, #16]

0x467344 UBFM	X7, X21, #61, #60

0x467348 ADD	X9, X30, X7

0x46734c LDR	Q20, [X30, X7]

0x467350 LDR	D21, [X9, #16]

0x467354 FMLA	V31.2D, V20.2D, V0.D[0]

0x467358 FMADD	D15, D21, D0, D15

0x46735c LDR	X0, [X19, #1096]

0x467360 STR	Q31, [SP, #144]

0x467364 BL	4c4660

0x467368 LDR	Q1, [SP, #144]

0x46736c FMOV	D2, D15

0x467370 LDP	X29, X30, [SP, #32]

0x467374 LDP	X19, X20, [SP, #48]

0x467378 STR	Q1, [SP, #240]

0x46737c LDR	D15, [SP, #128]

0x467380 LDP	X21, X22, [SP, #64]

0x467384 LDP	X25, X26, [SP, #96]

0x467388 LDP	D0, D1, [SP, #240]

0x46738c ADD	SP, SP, #272

0x467390 RET

0x467394 LDR	X3, [X19, #1040]

0x467398 CBZ	X3, 4673bc

0x46739c UBFM	X2, X3, #61, #60

0x4673a0 LDR	X0, [X19, #1056]

0x4673a4 CMP	X2, #8

0x4673a8 B.LE	4673e4

0x4673ac BL	403a40

0x4673b0 LDR	X3, [X19, #1040]

0x4673b4 LDR	X0, [X19, #1056]

0x4673b8 B	467198

0x4673bc MOVI	V31.2D, #0

0x4673c0 FMOV	D15, D31

0x4673c4 B	46735c

0x4673c8 MOVI	V31.2D, #0

0x4673cc MOVZ	W7, #0

0x4673d0 FMOV	D15, D31

0x4673d4 B	4672cc

0x4673e4 B.NE	467198

0x4673e8 LDR	D1, [X5, X7]

0x4673ec STR	D1, [X0]

0x4673f0 B	467198

/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsMatrix.h: 217 - 217

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

217:   inline Type_t* operator[](size_type i) { return X.data() + i * D2; }

/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp: 155 - 156

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

155:   for (int i = 0; i < n; i++)

156:     res += a[i] * b[i];

/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/QMCWaveFunctions/DiracDeterminantRef.cpp: 88 - 91

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

88:   GradType g = simd::dot(invRow.data(), dpsiM[WorkingIndex], invRow.size());

89:   RatioTimer->stop();

90:   return g;

91: }

/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsVector.h: 178 - 248

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

178:   inline size_t size() const { return nLocal; }

[...]

248:   inline pointer data() { return X; }

/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/Numerics/PETE/OperatorTags.h: 61 - 94

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

61:   inline typename BinaryReturn<T1, T2, OpMultiply>::Type_t operator()(const T1& a, const T2& b) const

62:   {

63:     return (a * b);

[...]

94:     (const_cast<T1&>(a) += b);

/home/hbollore/gnu/unknow_version/include/c++/14.0.1/bits/stl_algobase.h: 413 - 453

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

413: 	{ *__to = *__from; }

[...]

451: 	  if (__builtin_expect(_Num > 1, true))

452: 	    __builtin_memmove(__result, __first, sizeof(_Tp) * _Num);

453: 	  else if (_Num == 1)

/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/TinyVector.h: 62 - 62

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

62:       X[d] = T(0);

/home/hbollore/gnu/unknow_version/include/c++/14.0.1/bits/stl_algo.h: 683 - 709

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

683:     { return std::copy(__first, __first + __n, __result); }

[...]

709:       if (__n2 <= 0)

Coverage (%)	Name	Source Location	Module

Path /

Metric	Value
CQA speedup if no scalar integer	2.13
CQA speedup if FP arith vectorized	1.44
CQA speedup if fully vectorized	1.46
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.41
Bottlenecks	micro-operation queue,
Function	miniqmcreference::DiracDeterminantRef >::evalGrad(qmcplusplus::ParticleSet&, int)
Source	OhmmsMatrix.h:217-217,inner_product.hpp:155-156,DiracDeterminantRef.cpp:88-91,OhmmsVector.h:178-178,OhmmsVector.h:248-248,OperatorTags.h:61-63,OperatorTags.h:94-94,stl_algobase.h:413-413,stl_algobase.h:451-453,TinyVector.h:62-62,stl_algo.h:683-683,stl_algo.h:709-709
Source loop unroll info	NA
Source loop unroll confidence level	NA
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	15.25
CQA cycles if no scalar integer	7.17
CQA cycles if FP arith vectorized	10.56
CQA cycles if fully vectorized	10.45
Front-end cycles	15.25
DIV/SQRT cycles	8.50
P0 cycles	8.50
P1 cycles	10.75
P2 cycles	10.75
P3 cycles	10.75
P4 cycles	10.75
P5 cycles	9.00
P6 cycles	9.00
P7 cycles	9.00
P8 cycles	9.00
P9 cycles	10.83
P10 cycles	10.50
P11 cycles	10.67
P12 cycles	0.00
P13 cycles	0.00
P14 cycles	0.00
Inter-iter dependencies cycles	NA
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	122.00
Nb uops	122.00
Nb loads	NA
Nb stores	3.00
Nb stack references	0.00
FLOP/cycle	3.34
Nb FLOP add-sub	9.00
Nb FLOP mul	0.00
Nb FLOP fma	21.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	28.33
Bytes prefetched	0.00
Bytes loaded	392.00
Bytes stored	40.00
Stride 0	NA
Stride 1	NA
Stride n	NA
Stride unknown	NA
Stride indirect	NA
Vectorization ratio all	46.27
Vectorization ratio load	55.56
Vectorization ratio store	66.67
Vectorization ratio mul	NA
Vectorization ratio add_sub	21.43
Vectorization ratio fma	81.82
Vectorization ratio div_sqrt	NA
Vectorization ratio other	33.33
Vector-efficiency ratio all	37.13
Vector-efficiency ratio load	47.22
Vector-efficiency ratio store	41.67
Vector-efficiency ratio mul	NA
Vector-efficiency ratio add_sub	27.68
Vector-efficiency ratio fma	45.45
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	29.76

Metric	Value
CQA speedup if no scalar integer	2.13
CQA speedup if FP arith vectorized	1.44
CQA speedup if fully vectorized	1.46
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.41
Bottlenecks	micro-operation queue,
Function	miniqmcreference::DiracDeterminantRef >::evalGrad(qmcplusplus::ParticleSet&, int)
Source	OhmmsMatrix.h:217-217,inner_product.hpp:155-156,DiracDeterminantRef.cpp:88-91,OhmmsVector.h:178-178,OhmmsVector.h:248-248,OperatorTags.h:61-63,OperatorTags.h:94-94,stl_algobase.h:413-413,stl_algobase.h:451-453,TinyVector.h:62-62,stl_algo.h:683-683,stl_algo.h:709-709
Source loop unroll info	NA
Source loop unroll confidence level	NA
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	15.25
CQA cycles if no scalar integer	7.17
CQA cycles if FP arith vectorized	10.56
CQA cycles if fully vectorized	10.45
Front-end cycles	15.25
DIV/SQRT cycles	8.50
P0 cycles	8.50
P1 cycles	10.75
P2 cycles	10.75
P3 cycles	10.75
P4 cycles	10.75
P5 cycles	9.00
P6 cycles	9.00
P7 cycles	9.00
P8 cycles	9.00
P9 cycles	10.83
P10 cycles	10.50
P11 cycles	10.67
P12 cycles	0.00
P13 cycles	0.00
P14 cycles	0.00
Inter-iter dependencies cycles	NA
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	122.00
Nb uops	122.00
Nb loads	NA
Nb stores	3.00
Nb stack references	0.00
FLOP/cycle	3.34
Nb FLOP add-sub	9.00
Nb FLOP mul	0.00
Nb FLOP fma	21.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	28.33
Bytes prefetched	0.00
Bytes loaded	392.00
Bytes stored	40.00
Stride 0	NA
Stride 1	NA
Stride n	NA
Stride unknown	NA
Stride indirect	NA
Vectorization ratio all	46.27
Vectorization ratio load	55.56
Vectorization ratio store	66.67
Vectorization ratio mul	NA
Vectorization ratio add_sub	21.43
Vectorization ratio fma	81.82
Vectorization ratio div_sqrt	NA
Vectorization ratio other	33.33
Vector-efficiency ratio all	37.13
Vector-efficiency ratio load	47.22
Vector-efficiency ratio store	41.67
Vector-efficiency ratio mul	NA
Vector-efficiency ratio add_sub	27.68
Vector-efficiency ratio fma	45.45
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	29.76

Path /

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

Function	miniqmcreference::DiracDeterminantRef >::evalGrad(qmcplusplus::ParticleSet&, int)
Source file and lines	OhmmsVector.h:178-248
Module	exec

The loop is defined in:

/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsMatrix.h:217
/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp:155-156
/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/QMCWaveFunctions/DiracDeterminantRef.cpp:88-91
/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsVector.h:178,248
/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/Numerics/PETE/OperatorTags.h:61-63,94
/home/hbollore/gnu/unknow_version/include/c++/14.0.1/bits/stl_algobase.h:413,451-453
/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/TinyVector.h:62
/home/hbollore/gnu/unknow_version/include/c++/14.0.1/bits/stl_algo.h:683,709

Warnings:

Non-innermost loop: analyzing only self part (ignoring child loops).
Ignoring paths for analysis
Failed to get the number of paths
RecMII not computed since number of paths is unknown or > max_paths

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 15.25 to 7.17 cycles (2.13x speedup).

Workaround

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

Vectorization

Your loop is poorly vectorized. Only 37% of vector register length is used (average across all VPU instructions). By fully vectorizing your loop, you can lower the cost of an iteration from 15.25 to 10.45 cycles (1.46x speedup).

Details

46% of VPU instructions are used in vector version (process two or more data elements in vector registers):

55% of VPU loads are used in vector version.
66% of VPU stores are used in vector version.
21% of VPU addition or subtraction instructions are used in vector version.
81% of VPU fused multiply-add instructions are used in vector version.
33% of VPU instructions that are not load, store, addition, subtraction nor multiply instructions are used in vector version.

Since your execution units are vector units, only a fully vectorized loop can use their full power.

Workaround

Try another compiler or update/tune your current one
Remove inter-iterations dependences from your loop and make it unit-stride:
- If your arrays have 2 or more dimensions, check whether elements are accessed contiguously and, otherwise, try to permute loops accordingly: C storage order is row-major: for(i) for(j) a[j][i] = b[j][i]; (slow, non stride 1) => for(i) for(j) a[i][j] = b[i][j]; (fast, stride 1)
- If your loop streams arrays of structures (AoS), try to use structures of arrays instead (SoA): for(i) a[i].x = b[i].x; (slow, non stride 1) => for(i) a.x[i] = b.x[i]; (fast, stride 1)

Expensive FP math instructions/calls

Detected performance impact from expensive FP math instructions/calls. By removing/reexpressing them, you can lower the cost of an iteration from 15.25 to 1.00 cycles (15.25x speedup).

FMA

Detected 21 FMA (fused multiply-add) operations.

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

The binary loop is composed of 51 FP arithmetical operations:

30: addition or subtraction (21 inside FMA instructions)
21: multiply (all inside FMA instructions)

The binary loop is loading 392 bytes. The binary loop is storing 40 bytes.

Arithmetic intensity

Arithmetic intensity is 0.12 FP operations per loaded or stored byte.

General properties

nb instructions	122
loop length	488
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

15.25 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	8.50	8.50	10.75	10.75	10.75	10.75	9.00	9.00	9.00	9.00	10.83	10.50	10.67	0.00	0.00
cycles	8.50	8.50	10.75	10.75	10.75	10.75	9.00	9.00	9.00	9.00	10.83	10.50	10.67	0.00	0.00

Execution ports to units layout:

P0:
P1:
P2: ALU
P3: ALU
P4: ALU
P5: ALU
P6 (128 bits): VPU, FP store data, ALU, DIV/SQRT
P7 (128 bits): VPU, ALU, FP store data
P8 (128 bits): VPU, ALU, DIV/SQRT
P9 (128 bits): ALU, VPU
P10 (256 bits): store address, load
P11 (256 bits): store address, load
P12 (256 bits): load
P13 (64 bits): store data
P14 (64 bits): store data

Cycles executing div or sqrt instructions

Cycles summary

Front-end	15.25
Overall L1	15.25

Vectorization ratios

INT

all	34%
load	55%
store	66%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	0%
fma	0%
other	33%

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

INT+FP

all	46%
load	55%
store	66%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	21%
fma	81%
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	33%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 15.25 cycles. At this rate:

26% of peak load performance is reached (25.70 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
5% of peak store performance is reached (2.62 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the loop is: 467394

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	Latency	Recip. throughput
ORR W16, WZR, W3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR X17, [X19, #280]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
LDR X30, [X19, #320]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
MADD X21, X21, X17, XZR	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1
CMP W3, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.LE 4673bc <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x3fc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
SUB W1, W3, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ORR W5, WZR, W3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
CMP W1, #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.LS 4673c8 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x408>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADD X9, X21, X21,LSL #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
UBFM W12, W3, #2, #31	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
MOVZ W2, #96	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
MOVI V26.2D, #0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
ORR X10, XZR, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD X14, X30, X9,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
UMADDL X11, W12, W2, X14	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1
ORR V4.16B, V26.16B, V26.16B	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
ORR V22.16B, V26.16B, V26.16B	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
ORR V3.16B, V26.16B, V26.16B	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
ORR V19.16B, V26.16B, V26.16B	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
ORR V5.16B, V26.16B, V26.16B	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
SUB X6, X11, X14	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
TBZ W6, #5, 467230 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x270>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDP Q28, Q29, [X0]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.67
ADD X10, X0, #32	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDP Q2, Q25, [X14, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.67
LDP Q24, Q30, [X14, #48]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.67
LDR Q27, [X14, #80]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
LDR Q6, [X14], #96	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	6	0.33
FMLA V19.2D, V28.2D, V2.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMLA V3.2D, V25.2D, V28.D[1]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMLA V22.2D, V24.2D, V29.D[0]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMLA V4.2D, V29.2D, V30.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMLA V5.2D, V6.2D, V28.D[0]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMLA V26.2D, V27.2D, V29.D[1]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
CMP X14, X11	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.EQ 467298 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x2d8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
MOV D27, V3.D[1]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
AND W7, W5, #1886	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
EXT V3.16B, V19.16B, V3.16B, #8	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
MOV D6, V26.D[1]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
EXT V26.16B, V4.16B, V26.16B, #8	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
FADD V30.2D, V5.2D, V22.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
FADD D19, D4, D19	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
FADD V22.2D, V3.2D, V26.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
FADD D4, D27, D6	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
FADD V31.2D, V22.2D, V30.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
FADD D15, D4, D19	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
ANDS XZR, X5, #4160	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.EQ 46735c <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x39c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADD X15, X21, W7,SXTW	1	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	2	0.50
ADD X18, X15, X15,LSL #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ORR W25, WZR, W7	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
UBFM X20, X25, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD W4, W7, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
UBFM X8, X18, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD X22, X30, X8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR D5, [X0, X20]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
LDR Q7, [X30, X8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
LDR D16, [X22, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
FMLA V31.2D, V7.2D, V5.D[0]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMADD D15, D5, D16, D15	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
CMP W16, W4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.LE 46735c <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x39c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADD X26, X21, W4,SXTW	1	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	2	0.50
ADD X3, X26, X26,LSL #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD X17, X0, X20	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD W1, W7, #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
UBFM X0, X3, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR D23, [X17, #8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
ADD X5, X30, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR Q17, [X30, X0]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
LDR D18, [X5, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
FMLA V31.2D, V17.2D, V23.D[0]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMADD D15, D23, D18, D15	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
CMP W16, W1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.LE 46735c <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x39c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADD X16, X21, W1,SXTW	1	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	2	0.50
ADD X21, X16, X16,LSL #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR D0, [X17, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
UBFM X7, X21, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD X9, X30, X7	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR Q20, [X30, X7]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
LDR D21, [X9, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
FMLA V31.2D, V20.2D, V0.D[0]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMADD D15, D21, D0, D15	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
LDR X0, [X19, #1096]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
STR Q31, [SP, #144]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50
BL 4c4660 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE4stopEv>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	1	0.50
LDR Q1, [SP, #144]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
FMOV D2, D15	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
LDP X29, X30, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	1
LDP X19, X20, [SP, #48]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	1
STR Q1, [SP, #240]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50
LDR D15, [SP, #128]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
LDP X21, X22, [SP, #64]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	1
LDP X25, X26, [SP, #96]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	1
LDP D0, D1, [SP, #240]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
ADD SP, SP, #272	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
RET	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR X3, [X19, #1040]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
CBZ X3, 4673bc <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x3fc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
UBFM X2, X3, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR X0, [X19, #1056]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
CMP X2, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.LE 4673e4 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x424>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
BL 403a40 <@plt_start@+0x50>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	1	0.50
LDR X3, [X19, #1040]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
LDR X0, [X19, #1056]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
B 467198 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x1d8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
MOVI V31.2D, #0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
FMOV D15, D31	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
B 46735c <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x39c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
MOVI V31.2D, #0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
MOVZ W7, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
FMOV D15, D31	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
B 4672cc <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x30c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
B.NE 467198 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x1d8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR D1, [X5, X7]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
STR D1, [X0]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50
B 467198 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x1d8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50

Function	miniqmcreference::DiracDeterminantRef >::evalGrad(qmcplusplus::ParticleSet&, int)
Source file and lines	OhmmsVector.h:178-248
Module	exec

The loop is defined in:

/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsMatrix.h:217
/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp:155-156
/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/QMCWaveFunctions/DiracDeterminantRef.cpp:88-91
/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsVector.h:178,248
/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/Numerics/PETE/OperatorTags.h:61-63,94
/home/hbollore/gnu/unknow_version/include/c++/14.0.1/bits/stl_algobase.h:413,451-453
/home/hbollore/qaas-runs/171-275-7410/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/TinyVector.h:62
/home/hbollore/gnu/unknow_version/include/c++/14.0.1/bits/stl_algo.h:683,709

Warnings:

Non-innermost loop: analyzing only self part (ignoring child loops).
Ignoring paths for analysis
Failed to get the number of paths
RecMII not computed since number of paths is unknown or > max_paths

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 15.25 to 7.17 cycles (2.13x speedup).

Workaround

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

Vectorization

Details

46% of VPU instructions are used in vector version (process two or more data elements in vector registers):

55% of VPU loads are used in vector version.
66% of VPU stores are used in vector version.
21% of VPU addition or subtraction instructions are used in vector version.
81% of VPU fused multiply-add instructions are used in vector version.
33% of VPU instructions that are not load, store, addition, subtraction nor multiply instructions are used in vector version.

Since your execution units are vector units, only a fully vectorized loop can use their full power.

Workaround

Try another compiler or update/tune your current one
Remove inter-iterations dependences from your loop and make it unit-stride:
- If your arrays have 2 or more dimensions, check whether elements are accessed contiguously and, otherwise, try to permute loops accordingly: C storage order is row-major: for(i) for(j) a[j][i] = b[j][i]; (slow, non stride 1) => for(i) for(j) a[i][j] = b[i][j]; (fast, stride 1)
- If your loop streams arrays of structures (AoS), try to use structures of arrays instead (SoA): for(i) a[i].x = b[i].x; (slow, non stride 1) => for(i) a.x[i] = b.x[i]; (fast, stride 1)

Expensive FP math instructions/calls

Detected performance impact from expensive FP math instructions/calls. By removing/reexpressing them, you can lower the cost of an iteration from 15.25 to 1.00 cycles (15.25x speedup).

FMA

Detected 21 FMA (fused multiply-add) operations.

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

The binary loop is composed of 51 FP arithmetical operations:

30: addition or subtraction (21 inside FMA instructions)
21: multiply (all inside FMA instructions)

The binary loop is loading 392 bytes. The binary loop is storing 40 bytes.

Arithmetic intensity

Arithmetic intensity is 0.12 FP operations per loaded or stored byte.

General properties

nb instructions	122
loop length	488
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

15.25 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	8.50	8.50	10.75	10.75	10.75	10.75	9.00	9.00	9.00	9.00	10.83	10.50	10.67	0.00	0.00
cycles	8.50	8.50	10.75	10.75	10.75	10.75	9.00	9.00	9.00	9.00	10.83	10.50	10.67	0.00	0.00

Execution ports to units layout:

P0:
P1:
P2: ALU
P3: ALU
P4: ALU
P5: ALU
P6 (128 bits): VPU, FP store data, ALU, DIV/SQRT
P7 (128 bits): VPU, ALU, FP store data
P8 (128 bits): VPU, ALU, DIV/SQRT
P9 (128 bits): ALU, VPU
P10 (256 bits): store address, load
P11 (256 bits): store address, load
P12 (256 bits): load
P13 (64 bits): store data
P14 (64 bits): store data

Cycles executing div or sqrt instructions

Cycles summary

Front-end	15.25
Overall L1	15.25

Vectorization ratios

INT

all	34%
load	55%
store	66%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	0%
fma	0%
other	33%

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

INT+FP

all	46%
load	55%
store	66%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	21%
fma	81%
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	33%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 15.25 cycles. At this rate:

26% of peak load performance is reached (25.70 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
5% of peak store performance is reached (2.62 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the loop is: 467394

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	Latency	Recip. throughput
ORR W16, WZR, W3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR X17, [X19, #280]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
LDR X30, [X19, #320]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
MADD X21, X21, X17, XZR	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1
CMP W3, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.LE 4673bc <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x3fc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
SUB W1, W3, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ORR W5, WZR, W3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
CMP W1, #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.LS 4673c8 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x408>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADD X9, X21, X21,LSL #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
UBFM W12, W3, #2, #31	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
MOVZ W2, #96	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
MOVI V26.2D, #0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
ORR X10, XZR, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD X14, X30, X9,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
UMADDL X11, W12, W2, X14	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1
ORR V4.16B, V26.16B, V26.16B	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
ORR V22.16B, V26.16B, V26.16B	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
ORR V3.16B, V26.16B, V26.16B	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
ORR V19.16B, V26.16B, V26.16B	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
ORR V5.16B, V26.16B, V26.16B	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
SUB X6, X11, X14	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
TBZ W6, #5, 467230 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x270>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDP Q28, Q29, [X0]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.67
ADD X10, X0, #32	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDP Q2, Q25, [X14, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.67
LDP Q24, Q30, [X14, #48]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.67
LDR Q27, [X14, #80]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
LDR Q6, [X14], #96	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	6	0.33
FMLA V19.2D, V28.2D, V2.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMLA V3.2D, V25.2D, V28.D[1]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMLA V22.2D, V24.2D, V29.D[0]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMLA V4.2D, V29.2D, V30.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMLA V5.2D, V6.2D, V28.D[0]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMLA V26.2D, V27.2D, V29.D[1]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
CMP X14, X11	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.EQ 467298 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x2d8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
MOV D27, V3.D[1]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
AND W7, W5, #1886	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
EXT V3.16B, V19.16B, V3.16B, #8	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
MOV D6, V26.D[1]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
EXT V26.16B, V4.16B, V26.16B, #8	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
FADD V30.2D, V5.2D, V22.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
FADD D19, D4, D19	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
FADD V22.2D, V3.2D, V26.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
FADD D4, D27, D6	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
FADD V31.2D, V22.2D, V30.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
FADD D15, D4, D19	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
ANDS XZR, X5, #4160	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.EQ 46735c <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x39c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADD X15, X21, W7,SXTW	1	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	2	0.50
ADD X18, X15, X15,LSL #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ORR W25, WZR, W7	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
UBFM X20, X25, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD W4, W7, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
UBFM X8, X18, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD X22, X30, X8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR D5, [X0, X20]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
LDR Q7, [X30, X8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
LDR D16, [X22, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
FMLA V31.2D, V7.2D, V5.D[0]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMADD D15, D5, D16, D15	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
CMP W16, W4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.LE 46735c <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x39c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADD X26, X21, W4,SXTW	1	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	2	0.50
ADD X3, X26, X26,LSL #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD X17, X0, X20	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD W1, W7, #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
UBFM X0, X3, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR D23, [X17, #8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
ADD X5, X30, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR Q17, [X30, X0]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
LDR D18, [X5, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
FMLA V31.2D, V17.2D, V23.D[0]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMADD D15, D23, D18, D15	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
CMP W16, W1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.LE 46735c <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x39c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADD X16, X21, W1,SXTW	1	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	2	0.50
ADD X21, X16, X16,LSL #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR D0, [X17, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
UBFM X7, X21, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
ADD X9, X30, X7	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR Q20, [X30, X7]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
LDR D21, [X9, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
FMLA V31.2D, V20.2D, V0.D[0]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
FMADD D15, D21, D0, D15	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25
LDR X0, [X19, #1096]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
STR Q31, [SP, #144]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50
BL 4c4660 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE4stopEv>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	1	0.50
LDR Q1, [SP, #144]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
FMOV D2, D15	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
LDP X29, X30, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	1
LDP X19, X20, [SP, #48]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	1
STR Q1, [SP, #240]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50
LDR D15, [SP, #128]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
LDP X21, X22, [SP, #64]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	1
LDP X25, X26, [SP, #96]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	1
LDP D0, D1, [SP, #240]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
ADD SP, SP, #272	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
RET	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR X3, [X19, #1040]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
CBZ X3, 4673bc <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x3fc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
UBFM X2, X3, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
LDR X0, [X19, #1056]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
CMP X2, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33
B.LE 4673e4 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x424>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
BL 403a40 <@plt_start@+0x50>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	1	0.50
LDR X3, [X19, #1040]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
LDR X0, [X19, #1056]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	4	0.33
B 467198 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x1d8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
MOVI V31.2D, #0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
FMOV D15, D31	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
B 46735c <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x39c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
MOVI V31.2D, #0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
MOVZ W7, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25
FMOV D15, D31	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25
B 4672cc <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x30c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
B.NE 467198 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x1d8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR D1, [X5, X7]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33
STR D1, [X0]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50
B 467198 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE8evalGradERNS1_11ParticleSetEi+0x1d8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50

Report Configuration

Code clean check

Workaround

Vectorization

Details

Workaround

Expensive FP math instructions/calls

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

Workaround

Expensive FP math instructions/calls

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