OV - exec - Loop 1861

	Loop Id: 1861	Module: exec	Source: :0-0	Coverage: 4.44%

0x4c4270 VMOVSD	(%R14,%R9,8),%XMM23    [6]

0x4c4277 VSUBSD	%XMM20,%XMM23,%XMM23

0x4c427d VMOVSD	(%RSI,%R9,8),%XMM24    [5]

0x4c4284 VSUBSD	%XMM21,%XMM24,%XMM24

0x4c428a VMOVSD	(%R8,%R9,8),%XMM25    [1]

0x4c4291 VSUBSD	%XMM22,%XMM25,%XMM25

0x4c4297 VMULSD	%XMM23,%XMM2,%XMM26

0x4c429d VFMADD231SD	%XMM24,%XMM3,%XMM26

0x4c42a3 VFMADD231SD	%XMM25,%XMM4,%XMM26

0x4c42a9 VMULSD	%XMM23,%XMM5,%XMM27

0x4c42af VFMADD231SD	%XMM24,%XMM6,%XMM27

0x4c42b5 VFMADD231SD	%XMM25,%XMM7,%XMM27

0x4c42bb VMULSD	%XMM23,%XMM8,%XMM23

0x4c42c1 VFMADD231SD	%XMM24,%XMM9,%XMM23

0x4c42c7 VFMADD231SD	%XMM25,%XMM10,%XMM23

0x4c42cd VMOVDQA64	%XMM1,%XMM24

0x4c42d3 VPTERNLOGQ	$-0x8,%XMM0,%XMM26,%XMM24

0x4c42da VADDSD	%XMM24,%XMM26,%XMM24

0x4c42e0 VRNDSCALESD	$0xb,%XMM24,%XMM24,%XMM24

0x4c42e7 VSUBSD	%XMM24,%XMM26,%XMM24

0x4c42ed VMOVDQA64	%XMM1,%XMM25

0x4c42f3 VPTERNLOGQ	$-0x8,%XMM0,%XMM27,%XMM25

0x4c42fa VADDSD	%XMM25,%XMM27,%XMM25

0x4c4300 VRNDSCALESD	$0xb,%XMM25,%XMM25,%XMM25

0x4c4307 VSUBSD	%XMM25,%XMM27,%XMM25

0x4c430d VMOVDQA64	%XMM1,%XMM26

0x4c4313 VPTERNLOGQ	$-0x8,%XMM0,%XMM23,%XMM26

0x4c431a VADDSD	%XMM26,%XMM23,%XMM26

0x4c4320 VRNDSCALESD	$0xb,%XMM26,%XMM26,%XMM26

0x4c4327 VSUBSD	%XMM26,%XMM23,%XMM23

0x4c432d VMULSD	%XMM24,%XMM11,%XMM26

0x4c4333 VFMADD231SD	%XMM25,%XMM12,%XMM26

0x4c4339 VFMADD231SD	%XMM13,%XMM23,%XMM26

0x4c433f VMOVSD	%XMM26,(%RDX,%R9,8)    [7]

0x4c4346 VMULSD	%XMM24,%XMM14,%XMM27

0x4c434c VFMADD231SD	%XMM25,%XMM15,%XMM27

0x4c4352 VFMADD231SD	%XMM23,%XMM16,%XMM27

0x4c4358 VMOVSD	%XMM27,(%RDI,%R9,8)    [4]

0x4c435f VMULSD	%XMM24,%XMM17,%XMM24

0x4c4365 VFMADD231SD	%XMM25,%XMM18,%XMM24

0x4c436b VFMADD231SD	%XMM23,%XMM19,%XMM24

0x4c4371 VMOVSD	%XMM24,(%RCX,%R9,8)    [2]

0x4c4378 VMULSD	%XMM26,%XMM26,%XMM23

0x4c437e VFMADD231SD	%XMM27,%XMM27,%XMM23

0x4c4384 VFMADD231SD	%XMM24,%XMM24,%XMM23

0x4c438a VSQRTSD	%XMM23,%XMM23,%XMM23

0x4c4390 VMOVSD	%XMM23,(%R12,%R9,8)    [3]

0x4c4397 INC	%R9

0x4c439a CMP	%R9,%R13

0x4c439d JNE	4c4270

  ***  This Panel is Intentionally Left Blank.  ***

It is due to a lack of debug symbols in the given object
(loop or function).

Coverage (%)	Name	Module
►94.67+	qmcplusplus::ParticleSet::upda[...]	exec
○	qmcplusplus::NonLocalPP<double[...]	exec
○	main.extracted.109	exec
○	__kmp_invoke_microtask	libiomp5.so
○	__kmp_invoke_task_func	libiomp5.so
○	__kmp_launch_thread	libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]	libiomp5.so
○	start_thread	libc.so.6
○	__GI___clone3	libc.so.6
►2.85+	qmcplusplus::ParticleSet::upda[...]	exec
○	qmcplusplus::NonLocalPP<double[...]	exec
○	main.extracted.109	exec
○	__kmp_invoke_microtask	libiomp5.so
○	__kmp_invoke_task_func	libiomp5.so
○	__kmp_fork_call	libiomp5.so
○	__kmpc_fork_call	libiomp5.so
○	main	exec
○	__libc_start_call_main	libc.so.6
○	__libc_start_main	libc.so.6
○	_start	exec

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	4.29
CQA speedup if fully vectorized	8.00
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	NA
Bottlenecks	P10, P11,
Function	qmcplusplus::SoaDistanceTableABOMPTarget::evaluate(qmcplusplus::ParticleSet&)
Source
Source loop unroll info	NA
Source loop unroll confidence level	NA
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	11.00
CQA cycles if no scalar integer	11.00
CQA cycles if FP arith vectorized	2.56
CQA cycles if fully vectorized	1.38
Front-end cycles	6.13
P0 cycles	0.33
P1 cycles	0.33
P2 cycles	0.33
P3 cycles	0.33
P4 cycles	0.33
P5 cycles	0.33
P6 cycles	1.75
P7 cycles	1.75
P8 cycles	1.75
P9 cycles	1.75
P10 cycles	11.00
P11 cycles	11.00
P12 cycles	7.50
P13 cycles	7.50
P14 cycles	2.00
P15 cycles	2.00
DIV/SQRT cycles	8.50
Inter-iter dependencies cycles	1
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	50.00
Nb uops	49.00
Nb loads	3.00
Nb stores	4.00
Nb stack references	0.00
FLOP/cycle	4.09
Nb FLOP add-sub	9.00
Nb FLOP mul	7.00
Nb FLOP fma	14.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	1.00
Nb FLOP rsqrt	0.00
Bytes/cycle	5.09
Bytes prefetched	0.00
Bytes loaded	24.00
Bytes stored	32.00
Stride 0	0.00
Stride 1	7.00
Stride n	0.00
Stride unknown	0.00
Stride indirect	0.00
Vectorization ratio all	12.77
Vectorization ratio load	0.00
Vectorization ratio store	0.00
Vectorization ratio mul	0.00
Vectorization ratio add_sub	0.00
Vectorization ratio fma	0.00
Vectorization ratio div_sqrt	0.00
Vectorization ratio other	66.67
Vector-efficiency ratio all	14.10
Vector-efficiency ratio load	12.50
Vector-efficiency ratio store	12.50
Vector-efficiency ratio mul	12.50
Vector-efficiency ratio add_sub	12.50
Vector-efficiency ratio fma	12.50
Vector-efficiency ratio div_sqrt	12.50
Vector-efficiency ratio other	20.83

Metric	Value
CQA speedup if no scalar integer	1.00
CQA speedup if FP arith vectorized	4.29
CQA speedup if fully vectorized	8.00
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	NA
Bottlenecks	P10, P11,
Function	qmcplusplus::SoaDistanceTableABOMPTarget::evaluate(qmcplusplus::ParticleSet&)
Source
Source loop unroll info	NA
Source loop unroll confidence level	NA
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	11.00
CQA cycles if no scalar integer	11.00
CQA cycles if FP arith vectorized	2.56
CQA cycles if fully vectorized	1.38
Front-end cycles	6.13
P0 cycles	0.33
P1 cycles	0.33
P2 cycles	0.33
P3 cycles	0.33
P4 cycles	0.33
P5 cycles	0.33
P6 cycles	1.75
P7 cycles	1.75
P8 cycles	1.75
P9 cycles	1.75
P10 cycles	11.00
P11 cycles	11.00
P12 cycles	7.50
P13 cycles	7.50
P14 cycles	2.00
P15 cycles	2.00
DIV/SQRT cycles	8.50
Inter-iter dependencies cycles	1
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	50.00
Nb uops	49.00
Nb loads	3.00
Nb stores	4.00
Nb stack references	0.00
FLOP/cycle	4.09
Nb FLOP add-sub	9.00
Nb FLOP mul	7.00
Nb FLOP fma	14.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	1.00
Nb FLOP rsqrt	0.00
Bytes/cycle	5.09
Bytes prefetched	0.00
Bytes loaded	24.00
Bytes stored	32.00
Stride 0	0.00
Stride 1	7.00
Stride n	0.00
Stride unknown	0.00
Stride indirect	0.00
Vectorization ratio all	12.77
Vectorization ratio load	0.00
Vectorization ratio store	0.00
Vectorization ratio mul	0.00
Vectorization ratio add_sub	0.00
Vectorization ratio fma	0.00
Vectorization ratio div_sqrt	0.00
Vectorization ratio other	66.67
Vector-efficiency ratio all	14.10
Vector-efficiency ratio load	12.50
Vector-efficiency ratio store	12.50
Vector-efficiency ratio mul	12.50
Vector-efficiency ratio add_sub	12.50
Vector-efficiency ratio fma	12.50
Vector-efficiency ratio div_sqrt	12.50
Vector-efficiency ratio other	20.83

Path /

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

Function	qmcplusplus::SoaDistanceTableABOMPTarget::evaluate(qmcplusplus::ParticleSet&)
Source file and lines
Module	exec

gain
potential
hint
expert

Vectorization

Your loop is probably not vectorized. Only 14% of vector register length is used (average across all SSE/AVX instructions). By vectorizing your loop, you can lower the cost of an iteration from 11.00 to 1.37 cycles (8.00x speedup).

Details

Store and arithmetical SSE/AVX 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
- If your loop streams arrays of structures (AoS), try to use structures of arrays instead (SoA)

Execution units bottlenecks

Performance is limited by execution of FP multiply or FMA (fused multiply-add) operations (the FP multiply/FMA unit is a bottleneck).

Workaround

Reduce the number of FP multiply/FMA instructions

FMA

Detected 14 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).

Type of elements and instruction set

34 SSE or AVX instructions are processing arithmetic or math operations on double precision FP elements in scalar mode (one at a time).

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

The binary loop is composed of 45 FP arithmetical operations:

23: addition or subtraction (14 inside FMA instructions)
21: multiply (14 inside FMA instructions)
1: square root

The binary loop is loading 24 bytes (3 double precision FP elements). The binary loop is storing 32 bytes (4 double precision FP elements).

Arithmetic intensity

Arithmetic intensity is 0.80 FP operations per loaded or stored byte.

General properties

nb instructions	50
nb uops	49
loop length	307
used x86 registers	9
used mmx registers	0
used xmm registers	28
used ymm registers	0
used zmm registers	0
nb stack references	0
ADD-SUB / MUL ratio	1.29

Front-end

ASSUMED MACRO FUSION FIT IN UOP CACHE

micro-operation queue	6.13 cycles
front end	6.13 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15
uops	0.33	0.33	0.33	0.33	0.33	0.33	1.75	1.75	1.75	1.75	11.00	11.00	7.50	7.50	2.00	2.00
cycles	0.33	0.33	0.33	0.33	0.33	0.33	1.75	1.75	1.75	1.75	11.00	11.00	7.50	7.50	2.00	2.00

Execution ports to units layout:

P0: ALU
P1: ALU
P2: ALU
P3: ALU, BRU
P4: ALU, BRU
P5: ALU, BRU
P6 (512 bits): store address, load
P7 (512 bits): store address, load
P8 (512 bits): store address, load
P9 (512 bits): store address, load
P10 (512 bits): VPU, DIV/SQRT
P11 (512 bits): VPU, DIV/SQRT
P12 (512 bits): VPU
P13 (512 bits): VPU
P14 (512 bits): FP store data
P15 (512 bits): FP store data

Cycles executing div or sqrt instructions	8.50
Longest recurrence chain latency (RecMII)	1.00

Cycles summary

Front-end	6.13
Dispatch	11.00
DIV/SQRT	8.50
Data deps.	1.00
Overall L1	11.00

Vectorization ratios

INT

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	NA (no add-sub vectorizable/vectorized instructions)
fma	NA (no fma vectorizable/vectorized instructions)
other	100%

all	0%
load	0%
store	0%
mul	0%
add-sub	0%
fma	0%
div/sqrt	0%
other	0%

INT+FP

all	12%
load	0%
store	0%
mul	0%
add-sub	0%
fma	0%
div/sqrt	0%
other	66%

Vector efficiency ratios

INT

all	25%
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)
other	25%

all	12%
load	12%
store	12%
mul	12%
add-sub	12%
fma	12%
div/sqrt	12%
other	12%

INT+FP

all	14%
load	12%
store	12%
mul	12%
add-sub	12%
fma	12%
div/sqrt	12%
other	20%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 11.00 cycles. At this rate:

0% of peak load performance is reached (2.18 out of 256.00 bytes loaded per cycle (GB/s @ 1GHz))
2% of peak store performance is reached (2.91 out of 128.00 bytes stored per cycle (GB/s @ 1GHz))

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the loop is: 4c4270

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15	Latency	Recip. throughput	Vectorization
VMOVSD (%R14,%R9,8),%XMM23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	1	0.50	scal (12.5%)
VSUBSD %XMM20,%XMM23,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VMOVSD (%RSI,%R9,8),%XMM24	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	1	0.50	scal (12.5%)
VSUBSD %XMM21,%XMM24,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VMOVSD (%R8,%R9,8),%XMM25	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	1	0.50	scal (12.5%)
VSUBSD %XMM22,%XMM25,%XMM25	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VMULSD %XMM23,%XMM2,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (12.5%)
VFMADD231SD %XMM24,%XMM3,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VFMADD231SD %XMM25,%XMM4,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VMULSD %XMM23,%XMM5,%XMM27	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (12.5%)
VFMADD231SD %XMM24,%XMM6,%XMM27	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VFMADD231SD %XMM25,%XMM7,%XMM27	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VMULSD %XMM23,%XMM8,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (12.5%)
VFMADD231SD %XMM24,%XMM9,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VFMADD231SD %XMM25,%XMM10,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VMOVDQA64 %XMM1,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
VPTERNLOGQ $-0x8,%XMM0,%XMM26,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	1	0.25	vect (25.0%)
VADDSD %XMM24,%XMM26,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VRNDSCALESD $0xb,%XMM24,%XMM24,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	scal (12.5%)
VSUBSD %XMM24,%XMM26,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VMOVDQA64 %XMM1,%XMM25	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
VPTERNLOGQ $-0x8,%XMM0,%XMM27,%XMM25	1	0	0	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	1	0.25	vect (25.0%)
VADDSD %XMM25,%XMM27,%XMM25	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VRNDSCALESD $0xb,%XMM25,%XMM25,%XMM25	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	scal (12.5%)
VSUBSD %XMM25,%XMM27,%XMM25	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VMOVDQA64 %XMM1,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
VPTERNLOGQ $-0x8,%XMM0,%XMM23,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	1	0.25	vect (25.0%)
VADDSD %XMM26,%XMM23,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VRNDSCALESD $0xb,%XMM26,%XMM26,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	scal (12.5%)
VSUBSD %XMM26,%XMM23,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VMULSD %XMM24,%XMM11,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (12.5%)
VFMADD231SD %XMM25,%XMM12,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VFMADD231SD %XMM13,%XMM23,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VMOVSD %XMM26,(%RDX,%R9,8)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	1	0.50	scal (12.5%)
VMULSD %XMM24,%XMM14,%XMM27	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (12.5%)
VFMADD231SD %XMM25,%XMM15,%XMM27	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VFMADD231SD %XMM23,%XMM16,%XMM27	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VMOVSD %XMM27,(%RDI,%R9,8)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	1	0.50	scal (12.5%)
VMULSD %XMM24,%XMM17,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (12.5%)
VFMADD231SD %XMM25,%XMM18,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VFMADD231SD %XMM23,%XMM19,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VMOVSD %XMM24,(%RCX,%R9,8)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	1	0.50	scal (12.5%)
VMULSD %XMM26,%XMM26,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (12.5%)
VFMADD231SD %XMM27,%XMM27,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VFMADD231SD %XMM24,%XMM24,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VSQRTSD %XMM23,%XMM23,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	21	8.50	scal (12.5%)
VMOVSD %XMM23,(%R12,%R9,8)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	1	0.50	scal (12.5%)
INC %R9	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
CMP %R9,%R13	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
JNE 4c4270 <_ZN11qmcplusplus27SoaDistanceTableABOMPTargetIdLj3ELi40EE8evaluateERNS_11ParticleSetE+0x3e0>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A

Function	qmcplusplus::SoaDistanceTableABOMPTarget::evaluate(qmcplusplus::ParticleSet&)
Source file and lines
Module	exec

gain
potential
hint
expert

Vectorization

Details

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
- If your loop streams arrays of structures (AoS), try to use structures of arrays instead (SoA)

Execution units bottlenecks

Performance is limited by execution of FP multiply or FMA (fused multiply-add) operations (the FP multiply/FMA unit is a bottleneck).

Workaround

Reduce the number of FP multiply/FMA instructions

FMA

Detected 14 FMA (fused multiply-add) operations. Presence of both ADD/SUB and MUL operations.

Workaround

Type of elements and instruction set

34 SSE or AVX instructions are processing arithmetic or math operations on double precision FP elements in scalar mode (one at a time).

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

The binary loop is composed of 45 FP arithmetical operations:

23: addition or subtraction (14 inside FMA instructions)
21: multiply (14 inside FMA instructions)
1: square root

The binary loop is loading 24 bytes (3 double precision FP elements). The binary loop is storing 32 bytes (4 double precision FP elements).

Arithmetic intensity

Arithmetic intensity is 0.80 FP operations per loaded or stored byte.

General properties

nb instructions	50
nb uops	49
loop length	307
used x86 registers	9
used mmx registers	0
used xmm registers	28
used ymm registers	0
used zmm registers	0
nb stack references	0
ADD-SUB / MUL ratio	1.29

Front-end

ASSUMED MACRO FUSION FIT IN UOP CACHE

micro-operation queue	6.13 cycles
front end	6.13 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15
uops	0.33	0.33	0.33	0.33	0.33	0.33	1.75	1.75	1.75	1.75	11.00	11.00	7.50	7.50	2.00	2.00
cycles	0.33	0.33	0.33	0.33	0.33	0.33	1.75	1.75	1.75	1.75	11.00	11.00	7.50	7.50	2.00	2.00

Execution ports to units layout:

P0: ALU
P1: ALU
P2: ALU
P3: ALU, BRU
P4: ALU, BRU
P5: ALU, BRU
P6 (512 bits): store address, load
P7 (512 bits): store address, load
P8 (512 bits): store address, load
P9 (512 bits): store address, load
P10 (512 bits): VPU, DIV/SQRT
P11 (512 bits): VPU, DIV/SQRT
P12 (512 bits): VPU
P13 (512 bits): VPU
P14 (512 bits): FP store data
P15 (512 bits): FP store data

Cycles executing div or sqrt instructions	8.50
Longest recurrence chain latency (RecMII)	1.00

Cycles summary

Front-end	6.13
Dispatch	11.00
DIV/SQRT	8.50
Data deps.	1.00
Overall L1	11.00

Vectorization ratios

INT

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	NA (no add-sub vectorizable/vectorized instructions)
fma	NA (no fma vectorizable/vectorized instructions)
other	100%

all	0%
load	0%
store	0%
mul	0%
add-sub	0%
fma	0%
div/sqrt	0%
other	0%

INT+FP

all	12%
load	0%
store	0%
mul	0%
add-sub	0%
fma	0%
div/sqrt	0%
other	66%

Vector efficiency ratios

INT

all	25%
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)
other	25%

all	12%
load	12%
store	12%
mul	12%
add-sub	12%
fma	12%
div/sqrt	12%
other	12%

INT+FP

all	14%
load	12%
store	12%
mul	12%
add-sub	12%
fma	12%
div/sqrt	12%
other	20%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 11.00 cycles. At this rate:

0% of peak load performance is reached (2.18 out of 256.00 bytes loaded per cycle (GB/s @ 1GHz))
2% of peak store performance is reached (2.91 out of 128.00 bytes stored per cycle (GB/s @ 1GHz))

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the loop is: 4c4270

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15	Latency	Recip. throughput	Vectorization
VMOVSD (%R14,%R9,8),%XMM23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	1	0.50	scal (12.5%)
VSUBSD %XMM20,%XMM23,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VMOVSD (%RSI,%R9,8),%XMM24	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	1	0.50	scal (12.5%)
VSUBSD %XMM21,%XMM24,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VMOVSD (%R8,%R9,8),%XMM25	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	1	0.50	scal (12.5%)
VSUBSD %XMM22,%XMM25,%XMM25	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VMULSD %XMM23,%XMM2,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (12.5%)
VFMADD231SD %XMM24,%XMM3,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VFMADD231SD %XMM25,%XMM4,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VMULSD %XMM23,%XMM5,%XMM27	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (12.5%)
VFMADD231SD %XMM24,%XMM6,%XMM27	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VFMADD231SD %XMM25,%XMM7,%XMM27	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VMULSD %XMM23,%XMM8,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (12.5%)
VFMADD231SD %XMM24,%XMM9,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VFMADD231SD %XMM25,%XMM10,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VMOVDQA64 %XMM1,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
VPTERNLOGQ $-0x8,%XMM0,%XMM26,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	1	0.25	vect (25.0%)
VADDSD %XMM24,%XMM26,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VRNDSCALESD $0xb,%XMM24,%XMM24,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	scal (12.5%)
VSUBSD %XMM24,%XMM26,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VMOVDQA64 %XMM1,%XMM25	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
VPTERNLOGQ $-0x8,%XMM0,%XMM27,%XMM25	1	0	0	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	1	0.25	vect (25.0%)
VADDSD %XMM25,%XMM27,%XMM25	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VRNDSCALESD $0xb,%XMM25,%XMM25,%XMM25	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	scal (12.5%)
VSUBSD %XMM25,%XMM27,%XMM25	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VMOVDQA64 %XMM1,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
VPTERNLOGQ $-0x8,%XMM0,%XMM23,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	1	0.25	vect (25.0%)
VADDSD %XMM26,%XMM23,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VRNDSCALESD $0xb,%XMM26,%XMM26,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	scal (12.5%)
VSUBSD %XMM26,%XMM23,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	2	0.50	scal (12.5%)
VMULSD %XMM24,%XMM11,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (12.5%)
VFMADD231SD %XMM25,%XMM12,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VFMADD231SD %XMM13,%XMM23,%XMM26	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VMOVSD %XMM26,(%RDX,%R9,8)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	1	0.50	scal (12.5%)
VMULSD %XMM24,%XMM14,%XMM27	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (12.5%)
VFMADD231SD %XMM25,%XMM15,%XMM27	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VFMADD231SD %XMM23,%XMM16,%XMM27	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VMOVSD %XMM27,(%RDI,%R9,8)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	1	0.50	scal (12.5%)
VMULSD %XMM24,%XMM17,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (12.5%)
VFMADD231SD %XMM25,%XMM18,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VFMADD231SD %XMM23,%XMM19,%XMM24	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VMOVSD %XMM24,(%RCX,%R9,8)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	1	0.50	scal (12.5%)
VMULSD %XMM26,%XMM26,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	3	0.50	scal (12.5%)
VFMADD231SD %XMM27,%XMM27,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VFMADD231SD %XMM24,%XMM24,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	4	0.50	scal (12.5%)
VSQRTSD %XMM23,%XMM23,%XMM23	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	0	0	21	8.50	scal (12.5%)
VMOVSD %XMM23,(%R12,%R9,8)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	1	0.50	scal (12.5%)
INC %R9	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
CMP %R9,%R13	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
JNE 4c4270 <_ZN11qmcplusplus27SoaDistanceTableABOMPTargetIdLj3ELi40EE8evaluateERNS_11ParticleSetE+0x3e0>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A

Report Configuration

Vectorization

Details

Workaround

Execution units bottlenecks

Workaround

FMA

Workaround

Type of elements and instruction set

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

Arithmetic intensity

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code

Vectorization

Details

Workaround

Execution units bottlenecks

Workaround

FMA

Workaround

Type of elements and instruction set

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

Arithmetic intensity

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code