OV - exec - Loop 841

0x25b440 VMOVUPD	(%RCX,%R8,8),%ZMM4    [2]

0x25b447 VMOVUPD	0x40(%RCX,%R8,8),%ZMM5    [2]

0x25b44f VMOVUPD	0x80(%RCX,%R8,8),%ZMM6    [2]

0x25b457 VMOVUPD	0xc0(%RCX,%R8,8),%ZMM7    [2]

0x25b45f VFMADD231PD	(%RAX,%R8,8),%ZMM4,%ZMM0    [1]

0x25b466 VFMADD231PD	0x40(%RAX,%R8,8),%ZMM5,%ZMM1    [1]

0x25b46e VFMADD231PD	0x80(%RAX,%R8,8),%ZMM6,%ZMM2    [1]

0x25b476 VFMADD231PD	0xc0(%RAX,%R8,8),%ZMM7,%ZMM3    [1]

0x25b47e ADD	$0x20,%R8

0x25b482 CMP	%R8,%RSI

0x25b485 JNE	25b440

/home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp: 82 - 83

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

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

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

Coverage (%)	Name	Source Location	Module
►95.41+	miniqmcreference::DiracDetermi[...]	DiracDeterminantRef.cpp:217	exec
○	qmcplusplus::WaveFunction::eva[...]	stl_vector.h:919	exec
○	qmcplusplus::NonLocalPP<double[...]	NonLocalPP.hpp:135	exec
○	main.omp_outlined.64	NewTimer.h:249	exec
○	__kmp_invoke_microtask		libomp.so
○	__kmp_invoke_task_func		libomp.so
○	__kmp_launch_thread		libomp.so
○	__kmp_launch_worker(void*)		libomp.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►3.72+	miniqmcreference::DiracDetermi[...]	DiracDeterminantRef.cpp:217	exec
○	qmcplusplus::WaveFunction::eva[...]	stl_vector.h:919	exec
○	qmcplusplus::NonLocalPP<double[...]	NonLocalPP.hpp:135	exec
○	main.omp_outlined.64	NewTimer.h:249	exec
○	__kmp_invoke_microtask		libomp.so
○	__kmp_invoke_task_func		libomp.so
○	__kmp_fork_call		libomp.so
○	__kmpc_fork_call		libomp.so
○	main	miniqmc.cpp:411	exec
○	__libc_start_call_main		libc.so.6

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	1.00
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	2.00
Bottlenecks
Function	qmcplusplus::SPOSet::evaluateDetRatios(qmcplusplus::VirtualParticleSet const&, qmcplusplus::Vector >&, qmcplusplus::Vector > const&, std::vector >&)
Source	inner_product.hpp:82-83
Source loop unroll info	unrolled by 32
Source loop unroll confidence level	max
Unroll/vectorization loop type	main
Unroll factor	32
CQA cycles	4.00
CQA cycles if no scalar integer	4.00
CQA cycles if FP arith vectorized	4.00
CQA cycles if fully vectorized	4.00
Front-end cycles	1.25
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	2.00
P7 cycles	2.00
P8 cycles	2.00
P9 cycles	2.00
P10 cycles	2.00
P11 cycles	2.00
P12 cycles	0.00
P13 cycles	0.00
P14 cycles	0.00
P15 cycles	0.00
DIV/SQRT cycles	0.00
Inter-iter dependencies cycles	4
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	11.00
Nb uops	10.00
Nb loads	8.00
Nb stores	0.00
Nb stack references	0.00
FLOP/cycle	16.00
Nb FLOP add-sub	0.00
Nb FLOP mul	0.00
Nb FLOP fma	32.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	128.00
Bytes prefetched	0.00
Bytes loaded	512.00
Bytes stored	0.00
Stride 0	0.00
Stride 1	2.00
Stride n	0.00
Stride unknown	0.00
Stride indirect	0.00
Vectorization ratio all	100.00
Vectorization ratio load	100.00
Vectorization ratio store	NA
Vectorization ratio mul	NA
Vectorization ratio add_sub	NA
Vectorization ratio fma	100.00
Vectorization ratio div_sqrt	NA
Vectorization ratio other	NA
Vector-efficiency ratio all	100.00
Vector-efficiency ratio load	100.00
Vector-efficiency ratio store	NA
Vector-efficiency ratio mul	NA
Vector-efficiency ratio add_sub	NA
Vector-efficiency ratio fma	100.00
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	NA

Metric	Value
CQA speedup if no scalar integer	1.00
CQA speedup if FP arith vectorized	1.00
CQA speedup if fully vectorized	1.00
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	2.00
Bottlenecks
Function	qmcplusplus::SPOSet::evaluateDetRatios(qmcplusplus::VirtualParticleSet const&, qmcplusplus::Vector >&, qmcplusplus::Vector > const&, std::vector >&)
Source	inner_product.hpp:82-83
Source loop unroll info	unrolled by 32
Source loop unroll confidence level	max
Unroll/vectorization loop type	main
Unroll factor	32
CQA cycles	4.00
CQA cycles if no scalar integer	4.00
CQA cycles if FP arith vectorized	4.00
CQA cycles if fully vectorized	4.00
Front-end cycles	1.25
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	2.00
P7 cycles	2.00
P8 cycles	2.00
P9 cycles	2.00
P10 cycles	2.00
P11 cycles	2.00
P12 cycles	0.00
P13 cycles	0.00
P14 cycles	0.00
P15 cycles	0.00
DIV/SQRT cycles	0.00
Inter-iter dependencies cycles	4
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	11.00
Nb uops	10.00
Nb loads	8.00
Nb stores	0.00
Nb stack references	0.00
FLOP/cycle	16.00
Nb FLOP add-sub	0.00
Nb FLOP mul	0.00
Nb FLOP fma	32.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	128.00
Bytes prefetched	0.00
Bytes loaded	512.00
Bytes stored	0.00
Stride 0	0.00
Stride 1	2.00
Stride n	0.00
Stride unknown	0.00
Stride indirect	0.00
Vectorization ratio all	100.00
Vectorization ratio load	100.00
Vectorization ratio store	NA
Vectorization ratio mul	NA
Vectorization ratio add_sub	NA
Vectorization ratio fma	100.00
Vectorization ratio div_sqrt	NA
Vectorization ratio other	NA
Vector-efficiency ratio all	100.00
Vector-efficiency ratio load	100.00
Vector-efficiency ratio store	NA
Vector-efficiency ratio mul	NA
Vector-efficiency ratio add_sub	NA
Vector-efficiency ratio fma	100.00
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	NA

Path /

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

Function	qmcplusplus::SPOSet::evaluateDetRatios(qmcplusplus::VirtualParticleSet const&, qmcplusplus::Vector >&, qmcplusplus::Vector > const&, std::vector >&)
Source file and lines	inner_product.hpp:82-83
Module	exec

The loop is defined in /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp:82-83.

It is main loop of related source loop which is unrolled by 32 (including vectorization).

gain
potential
hint
expert

Vectorization

Your loop is fully vectorized, using full register length.

Details

All SSE/AVX instructions are used in vector version (process two or more data elements in vector registers).

Execution units bottlenecks

Found no such bottlenecks but see expert reports for more complex bottlenecks.

FMA

Detected 32 FMA (fused multiply-add) operations.

Vector unaligned load/store instructions

Detected 4 optimal vector unaligned load/store instructions.

Details

VMOVUPD: 4 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp:83
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp:83
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp:83
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp:83

Workaround

Use vector aligned instructions:

align your arrays on 64 bytes boundaries: replace { void *p = malloc (size); } with { void *p; posix_memalign (&p, 64, size); }.
inform your compiler that your arrays are vector aligned: if array 'foo' is 64 bytes-aligned, define a pointer 'p_foo' as __builtin_assume_aligned (foo, 64) and use it instead of 'foo' in the loop.

Type of elements and instruction set

4 AVX-512 instructions are processing arithmetic or math operations on double precision FP elements in vector mode (eight at a time).

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

The binary loop is composed of 64 FP arithmetical operations:

32: addition or subtraction (all inside FMA instructions)
32: multiply (all inside FMA instructions)

The binary loop is loading 512 bytes (64 double precision FP elements).

Arithmetic intensity

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

General properties

nb instructions	11
nb uops	10
loop length	71
used x86 registers	4
used mmx registers	0
used xmm registers	0
used ymm registers	0
used zmm registers	8
nb stack references	0

Front-end

ASSUMED MACRO FUSION FIT IN UOP CACHE

micro-operation queue	1.25 cycles
front end	1.25 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	2.00	2.00	2.00	2.00	2.00	2.00	0.00	0.00	0.00	0.00
cycles	0.33	0.33	0.33	0.33	0.33	0.33	2.00	2.00	2.00	2.00	2.00	2.00	0.00	0.00	0.00	0.00

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	NA
Longest recurrence chain latency (RecMII)	4.00

Cycles summary

Front-end	1.25
Dispatch	2.00
Data deps.	4.00
Overall L1	4.00

Vectorization ratios

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

Vector efficiency ratios

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

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 4.00 cycles. At this rate:

50% of peak load performance is reached (128.00 out of 256.00 bytes loaded per cycle (GB/s @ 1GHz))

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the loop is: 25b440

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	Latency	Recip. throughput	Vectorization
VMOVUPD (%RCX,%R8,8),%ZMM4	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.50	vect (100.0%)
VMOVUPD 0x40(%RCX,%R8,8),%ZMM5	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.50	vect (100.0%)
VMOVUPD 0x80(%RCX,%R8,8),%ZMM6	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.50	vect (100.0%)
VMOVUPD 0xc0(%RCX,%R8,8),%ZMM7	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.50	vect (100.0%)
VFMADD231PD (%RAX,%R8,8),%ZMM4,%ZMM0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.50	0.50	4	0.50	vect (100.0%)
VFMADD231PD 0x40(%RAX,%R8,8),%ZMM5,%ZMM1	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.50	0.50	4	0.50	vect (100.0%)
VFMADD231PD 0x80(%RAX,%R8,8),%ZMM6,%ZMM2	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.50	0.50	4	0.50	vect (100.0%)
VFMADD231PD 0xc0(%RAX,%R8,8),%ZMM7,%ZMM3	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.50	0.50	4	0.50	vect (100.0%)
ADD $0x20,%R8	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
CMP %R8,%RSI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
JNE 25b440 <_ZN11qmcplusplus6SPOSet17evaluateDetRatiosERKNS_18VirtualParticleSetERNS_6VectorIdSaIdEEERKS6_RSt6vectorIdS5_E+0xc0>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A

Function	qmcplusplus::SPOSet::evaluateDetRatios(qmcplusplus::VirtualParticleSet const&, qmcplusplus::Vector >&, qmcplusplus::Vector > const&, std::vector >&)
Source file and lines	inner_product.hpp:82-83
Module	exec

The loop is defined in /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp:82-83.

It is main loop of related source loop which is unrolled by 32 (including vectorization).

gain
potential
hint
expert

Vectorization

Your loop is fully vectorized, using full register length.

Details

All SSE/AVX instructions are used in vector version (process two or more data elements in vector registers).

Execution units bottlenecks

Found no such bottlenecks but see expert reports for more complex bottlenecks.

FMA

Detected 32 FMA (fused multiply-add) operations.

Vector unaligned load/store instructions

Detected 4 optimal vector unaligned load/store instructions.

Details

VMOVUPD: 4 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp:83
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp:83
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp:83
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp:83

Workaround

Use vector aligned instructions:

align your arrays on 64 bytes boundaries: replace { void *p = malloc (size); } with { void *p; posix_memalign (&p, 64, size); }.
inform your compiler that your arrays are vector aligned: if array 'foo' is 64 bytes-aligned, define a pointer 'p_foo' as __builtin_assume_aligned (foo, 64) and use it instead of 'foo' in the loop.

Type of elements and instruction set

4 AVX-512 instructions are processing arithmetic or math operations on double precision FP elements in vector mode (eight at a time).

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

The binary loop is composed of 64 FP arithmetical operations:

32: addition or subtraction (all inside FMA instructions)
32: multiply (all inside FMA instructions)

The binary loop is loading 512 bytes (64 double precision FP elements).

Arithmetic intensity

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

General properties

nb instructions	11
nb uops	10
loop length	71
used x86 registers	4
used mmx registers	0
used xmm registers	0
used ymm registers	0
used zmm registers	8
nb stack references	0

Front-end

ASSUMED MACRO FUSION FIT IN UOP CACHE

micro-operation queue	1.25 cycles
front end	1.25 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	2.00	2.00	2.00	2.00	2.00	2.00	0.00	0.00	0.00	0.00
cycles	0.33	0.33	0.33	0.33	0.33	0.33	2.00	2.00	2.00	2.00	2.00	2.00	0.00	0.00	0.00	0.00

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	NA
Longest recurrence chain latency (RecMII)	4.00

Cycles summary

Front-end	1.25
Dispatch	2.00
Data deps.	4.00
Overall L1	4.00

Vectorization ratios

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

Vector efficiency ratios

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

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 4.00 cycles. At this rate:

50% of peak load performance is reached (128.00 out of 256.00 bytes loaded per cycle (GB/s @ 1GHz))

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the loop is: 25b440

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	Latency	Recip. throughput	Vectorization
VMOVUPD (%RCX,%R8,8),%ZMM4	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.50	vect (100.0%)
VMOVUPD 0x40(%RCX,%R8,8),%ZMM5	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.50	vect (100.0%)
VMOVUPD 0x80(%RCX,%R8,8),%ZMM6	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.50	vect (100.0%)
VMOVUPD 0xc0(%RCX,%R8,8),%ZMM7	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.50	vect (100.0%)
VFMADD231PD (%RAX,%R8,8),%ZMM4,%ZMM0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.50	0.50	4	0.50	vect (100.0%)
VFMADD231PD 0x40(%RAX,%R8,8),%ZMM5,%ZMM1	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.50	0.50	4	0.50	vect (100.0%)
VFMADD231PD 0x80(%RAX,%R8,8),%ZMM6,%ZMM2	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.50	0.50	4	0.50	vect (100.0%)
VFMADD231PD 0xc0(%RAX,%R8,8),%ZMM7,%ZMM3	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.50	0.50	4	0.50	vect (100.0%)
ADD $0x20,%R8	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
CMP %R8,%RSI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
JNE 25b440 <_ZN11qmcplusplus6SPOSet17evaluateDetRatiosERKNS_18VirtualParticleSetERNS_6VectorIdSaIdEEERKS6_RSt6vectorIdS5_E+0xc0>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A

Report Configuration

Vectorization

Details

Execution units bottlenecks

FMA

Vector unaligned load/store instructions

Details

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

Execution units bottlenecks

FMA

Vector unaligned load/store instructions

Details

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