OV - exec - Function miniqmcreference::DiracDeterminantRef<qmcplusplus::DelayedUpdate<double, double> >::evaluateLog(qmcplusplus::ParticleSet&, qmcplusplus::ParticleAttrib<qmcplusplus::TinyVector<double, 3u>, std::allocator<qmcplusplus::TinyVector<double, 3u> > >&, qmcplusplus::ParticleAttrib<double, std::allocator<double> >&)

	Function: miniqmcreference::DiracDeterminantRef<qmcplusplus::DelayedUpdate<double, double> >::evalua ...	Module: exec	Source: DiracDeterminantRef.cpp:235-255 [...]	Coverage: 0.81%

/home/hbollore/qaas/qaas-runs/169-816-7419/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp: 81 - 154

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

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

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

[...]

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

/home/hbollore/qaas/qaas-runs/169-816-7419/intel/miniqmc/build/miniqmc/src/Numerics/PETE/OperatorTags.h: 63 - 94

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

63:     return (a * b);

[...]

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

/home/hbollore/qaas/qaas-runs/169-816-7419/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsVector.h: 223 - 223

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

223:     return X[i];

/home/hbollore/qaas/qaas-runs/169-816-7419/intel/miniqmc/build/miniqmc/src/QMCWaveFunctions/DiracDeterminantRef.cpp: 235 - 255

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

235: {

236:   recompute(P);

237:

238:   if (NumPtcls == 1)

239:   {

240:     ValueType y = psiM(0, 0);

241:     GradType rv = y * dpsiM(0, 0);

242:     G[FirstIndex] += rv;

243:     L[FirstIndex] += y * d2psiM(0, 0) - dot(rv, rv);

244:   }

245:   else

246:   {

247:     for (int i = 0, iat = FirstIndex; i < NumPtcls; i++, iat++)

248:     {

249:       mGradType rv   = simd::dot(psiM[i], dpsiM[i], NumOrbitals);

250:       mValueType lap = simd::dot(psiM[i], d2psiM[i], NumOrbitals);

251:       G[iat] += rv;

252:       L[iat] += lap - dot(rv, rv);

253:     }

254:   }

255:   return LogValue;

0x439700 STP	X29, X30, [SP, #960]!

0x439704 STP	X24, X23, [SP, #16]

0x439708 STP	X22, X21, [SP, #32]

0x43970c STP	X20, X19, [SP, #48]

0x439710 ADD	X29, SP, #0

0x439714 ORR	X20, XZR, X3

0x439718 ORR	X21, XZR, X2

0x43971c ORR	X19, XZR, X0

0x439720 BL	43aec0

0x439724 LDR	W8, [X19, #1156]

0x439728 CMP	W8, #1

0x43972c B.NE	4397a0

0x439730 LDR	X8, [X19, #256]

0x439734 LDRSW	X9, [X19, #1144]

0x439738 LDR	X10, [X21, #24]

0x43973c MOVZ	W11, #24

0x439740 SMADDL	X10, W9, W11, X10

0x439744 LDR	D0, [X8]

0x439748 LDR	X8, [X19, #320]

0x43974c LDR	Q3, [X10]

0x439750 LDR	Q2, [X8]

0x439754 LDR	D1, [X8, #16]

0x439758 LDR	X8, [X19, #384]

0x43975c FMUL	V2.2D, V2.2D, V0.D[0]

0x439760 FMUL	D1, D1, D0

0x439764 FADD	V3.2D, V3.2D, V2.2D

0x439768 FMUL	V2.2D, V2.2D, V2.2D

0x43976c STR	Q3, [X10]

0x439770 LDR	D3, [X10, #16]

0x439774 FADDP	D2, V2.2D

0x439778 FADD	D3, D3, D1

0x43977c FMADD	D1, D1, D1, D2

0x439780 STR	D3, [X10, #16]

0x439784 LDR	D3, [X8]

0x439788 LDR	X8, [X20, #24]

0x43978c LDR	D4, [X8, X9,LSL #3]

0x439790 FNMSUB	D0, D3, D0, D1

0x439794 FADD	D0, D0, D4

0x439798 STR	D0, [X8, X9,LSL #3]

0x43979c B	4399fc

0x4397a0 CMP	W8, #1

0x4397a4 B.LT	4399fc

0x4397a8 LDR	W9, [X19, #1152]

0x4397ac CMP	W9, #1

0x4397b0 B.LT	4399fc

0x4397b4 LDR	X17, [X19, #280]

0x4397b8 LDR	X1, [X19, #216]

0x4397bc LDR	X4, [X19, #344]

0x4397c0 ORR	X10, XZR, XZR

0x4397c4 CNTW	X16, ALL

0x4397c8 RDVL	X18, #3

0x4397cc LDR	X11, [X19, #256]

0x4397d0 LDR	X12, [X19, #320]

0x4397d4 RDVL	X3, #2

0x4397d8 LDR	X13, [X19, #384]

0x4397dc LDR	X14, [X21, #24]

0x4397e0 DUP	Z0.D, #0

0x4397e4 RDVL	X6, #1

0x4397e8 LDR	X15, [X20, #24]

0x4397ec PTRUE	P0.D, ALL

0x4397f0 LDRSW	X0, [X19, #1144]

0x4397f4 MOVZ	W5, #24

0x4397f8 UBFM	X1, X1, #61, #60

0x4397fc ADD	X2, X17, X17,LSL #1

0x439800 RDVL	X17, #6

0x439804 PTRUE	P1.B, ALL

0x439808 UBFM	X4, X4, #61, #60

0x43980c UBFM	X2, X2, #61, #60

0x439810 B	43987c

0x439814 HINT	#0

0x439818 HINT	#0

0x43981c HINT	#0

  (1048) 0x439820 MADD	X7, X0, X5, X14

  (1048) 0x439824 ADD	X10, X10, #1

  (1048) 0x439828 ADD	X11, X11, X1

  (1048) 0x43982c LDP	D5, D6, [X7]

  (1048) 0x439830 ADD	X12, X12, X2

  (1048) 0x439834 ADD	X13, X13, X4

  (1048) 0x439838 CMP	X10, X8

  (1048) 0x43983c FADD	D5, D5, D1

  (1048) 0x439840 STR	D5, [X7]

  (1048) 0x439844 FADD	D5, D6, D2

  (1048) 0x439848 STR	D5, [X7, #8]

  (1048) 0x43984c LDR	D5, [X7, #16]

  (1048) 0x439850 FADD	D5, D5, D3

  (1048) 0x439854 FMUL	D3, D3, D3

  (1048) 0x439858 STR	D5, [X7, #16]

  (1048) 0x43985c FMADD	D2, D2, D2, D3

  (1048) 0x439860 LDR	D5, [X15, X0,LSL #3]

  (1048) 0x439864 FMADD	D1, D1, D1, D2

  (1048) 0x439868 FSUB	D1, D4, S1

  (1048) 0x43986c FADD	D1, D1, D5

  (1048) 0x439870 STR	D1, [X15, X0,LSL #3]

  (1048) 0x439874 ADD	X0, X0, #1

  (1048) 0x439878 B.EQ	4399fc

  (1048) 0x43987c CMP	X16, X9

  (1048) 0x439880 B.LS	4398a0

  (1048) 0x439884 ORR	X7, XZR, XZR

  (1048) 0x439888 MOVI	D3, #0

  (1048) 0x43988c MOVI	D2, #0

  (1048) 0x439890 MOVI	D1, #0

  (1048) 0x439894 B	439938

0x439898 HINT	#0

0x43989c HINT	#0

  (1048) 0x4398a0 UDIV	X7, X9, X16

  (1048) 0x4398a4 ORR	X21, XZR, XZR

  (1048) 0x4398a8 ADDVL	X22, X11, #1

  (1048) 0x4398ac ORR	X23, XZR, X12

  (1048) 0x4398b0 ORR	Z3.D, Z0.D, Z0.D

  (1048) 0x4398b4 ORR	Z4.D, Z0.D, Z0.D

  (1048) 0x4398b8 ORR	Z2.D, Z0.D, Z0.D

  (1048) 0x4398bc ORR	Z5.D, Z0.D, Z0.D

  (1048) 0x4398c0 ORR	Z1.D, Z0.D, Z0.D

  (1048) 0x4398c4 ORR	Z6.D, Z0.D, Z0.D

  (1048) 0x4398c8 UMSUBL	X20, W7, W16, X9

  (1048) 0x4398cc UMADDL	X7, W7, W16, XZR

  (1048) 0x4398d0 HINT	#0

  (1048) 0x4398d4 HINT	#0

  (1048) 0x4398d8 HINT	#0

  (1048) 0x4398dc HINT	#0

    (1046) 0x4398e0 ADD	X24, X23, X18

    (1046) 0x4398e4 LD1D	{Z7.D}, P0/Z, [X11, X21,LSL #3]

    (1046) 0x4398e8 LD1D	{Z16.D}, P0/Z, [X22, X21,LSL #3]

    (1046) 0x4398ec LD3D	{Z17.D, Z18.D, Z19.D}, P0/Z, [X23, MUL VL]

    (1046) 0x4398f0 ADD	X21, X21, X16

    (1046) 0x4398f4 ADD	X23, X23, X17

    (1046) 0x4398f8 LD3D	{Z20.D, Z21.D, Z22.D}, P0/Z, [X24, MUL VL]

    (1046) 0x4398fc CMP	X7, X21

    (1046) 0x439900 FMLA	Z1.D, P0/M, Z17.D, Z7.D

    (1046) 0x439904 FMLA	Z2.D, P0/M, Z18.D, Z7.D

    (1046) 0x439908 FMLA	Z3.D, P0/M, Z19.D, Z7.D

    (1046) 0x43990c FMLA	Z6.D, P0/M, Z20.D, Z16.D

    (1046) 0x439910 FMLA	Z5.D, P0/M, Z21.D, Z16.D

    (1046) 0x439914 FMLA	Z4.D, P0/M, Z22.D, Z16.D

    (1046) 0x439918 B.NE	4398e0

  (1048) 0x43991c FADD	Z1.D, Z6.D, Z1.D

  (1048) 0x439920 FADD	Z2.D, Z5.D, Z2.D

  (1048) 0x439924 FADD	Z3.D, Z4.D, Z3.D

  (1048) 0x439928 FADDV	D1, P0, Z1.D

  (1048) 0x43992c FADDV	D2, P0, Z2.D

  (1048) 0x439930 FADDV	D3, P0, Z3.D

  (1048) 0x439934 CBZ	X20, 439968

  (1048) 0x439938 MADD	X20, X7, X5, X12

  (1048) 0x43993c HINT	#0

    (1049) 0x439940 LDR	D4, [X11, X7,LSL #3]

    (1049) 0x439944 LDP	D5, D6, [X20]

    (1049) 0x439948 LDR	D7, [X20, #16]

    (1049) 0x43994c ADD	X7, X7, #1

    (1049) 0x439950 ADD	X20, X20, #24

    (1049) 0x439954 CMP	X9, X7

    (1049) 0x439958 FMADD	D1, D5, D4, D1

    (1049) 0x43995c FMADD	D2, D6, D4, D2

    (1049) 0x439960 FMADD	D3, D7, D4, D3

    (1049) 0x439964 B.NE	439940

  (1048) 0x439968 CMP	X16, X9

  (1048) 0x43996c B.LS	439980

(1045) 0x439970 ORR	X7, XZR, XZR

(1045) 0x439974 MOVI	D4, #0

(1045) 0x439978 B	4399e0

0x43997c HINT	#0

  (1048) 0x439980 UDIV	X7, X9, X16

  (1048) 0x439984 ORR	X22, XZR, X13

  (1048) 0x439988 ORR	X23, XZR, X11

  (1048) 0x43998c ORR	Z4.D, Z0.D, Z0.D

  (1048) 0x439990 ORR	Z5.D, Z0.D, Z0.D

  (1048) 0x439994 UMSUBL	X20, W7, W16, X9

  (1048) 0x439998 UMADDL	X7, W7, W16, XZR

  (1048) 0x43999c ORR	X21, XZR, X7

    (1047) 0x4399a0 LD1D	{Z6.D}, P0/Z, [X23, MUL VL]

    (1047) 0x4399a4 LD1B	{Z7.B}, P1/Z, [X23, X6]

    (1047) 0x4399a8 LD1D	{Z16.D}, P0/Z, [X22, MUL VL]

    (1047) 0x4399ac LD1B	{Z17.B}, P1/Z, [X22, X6]

    (1047) 0x4399b0 ADD	X23, X23, X3

    (1047) 0x4399b4 ADD	X22, X22, X3

    (1047) 0x4399b8 SUBS	X21, X21, X16

    (1047) 0x4399bc FMLA	Z4.D, P0/M, Z16.D, Z6.D

    (1047) 0x4399c0 FMLA	Z5.D, P0/M, Z17.D, Z7.D

    (1047) 0x4399c4 B.NE	4399a0

  (1048) 0x4399c8 FADD	Z4.D, Z5.D, Z4.D

  (1048) 0x4399cc FADDV	D4, P0, Z4.D

  (1048) 0x4399d0 CBZ	X20, 439820

0x4399d4 HINT	#0

0x4399d8 HINT	#0

0x4399dc HINT	#0

(1045) 0x4399e0 LDR	D5, [X11, X7,LSL #3]

(1045) 0x4399e4 LDR	D6, [X13, X7,LSL #3]

(1045) 0x4399e8 ADD	X7, X7, #1

(1045) 0x4399ec CMP	X9, X7

(1045) 0x4399f0 FMADD	D4, D6, D5, D4

(1045) 0x4399f4 B.NE	4399e0

(1045) 0x4399f8 B	439820

0x4399fc LDR	D0, [X19, #16]

0x439a00 LDP	X20, X19, [SP, #48]

0x439a04 LDP	X22, X21, [SP, #32]

0x439a08 LDP	X24, X23, [SP, #16]

0x439a0c LDP	X29, X30, [SP], #64

0x439a10 RET

0x439a14 HINT	#0

0x439a18 HINT	#0

0x439a1c HINT	#0

Coverage (%)	Name	Source Location	Module
►100.00+	qmcplusplus::WaveFunction::eva[...]	WaveFunction.cpp:177	exec
○	.omp_outlined.	miniqmc.cpp:375	exec
○	__kmp_invoke_microtask		libomp.so

Path /

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

Source file and lines	DiracDeterminantRef.cpp:235-255
Module	exec

The function is defined in:

/home/hbollore/qaas/qaas-runs/169-816-7419/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp:81-82,154
/home/hbollore/qaas/qaas-runs/169-816-7419/intel/miniqmc/build/miniqmc/src/Numerics/PETE/OperatorTags.h:63,94
/home/hbollore/qaas/qaas-runs/169-816-7419/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsVector.h:223
/home/hbollore/qaas/qaas-runs/169-816-7419/intel/miniqmc/build/miniqmc/src/QMCWaveFunctions/DiracDeterminantRef.cpp:235-255

Warnings:
Ignoring paths for analysis

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 12.17 to 4.17 cycles (2.92x speedup).

Workaround

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

Vectorization

Your function is poorly vectorized. Only 33% of vector register length is used (average across all VPU instructions). By fully vectorizing your function, you can lower the cost of an iteration from 12.17 to 9.50 cycles (1.28x speedup).

Details

25% of VPU instructions are used in vector version (process two or more data elements in vector registers):

25% of VPU loads are used in vector version.
33% of VPU stores are used in vector version.
50% of VPU addition or subtraction instructions are used in vector version.
0% 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 function can use their full power.

Workaround

Try another compiler or update/tune your current one
Make array accesses unit-stride:
- If your function 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 12.17 to 0.50 cycles (24.33x speedup).

FMA

Detected 2 FMA (fused multiply-add) operations. Presence of both ADD/SUB and MUL operations.

Workaround

Try to change order in which elements are evaluated (using parentheses) in arithmetic expressions containing both ADD/SUB and MUL operations to enable your compiler to generate FMA instructions wherever possible. For instance a + b*c is a valid FMA (MUL then ADD). However (a+b)* c cannot be translated into an FMA (ADD then MUL).

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

The binary function is composed of 14 FP arithmetical operations:

7: addition or subtraction (2 inside FMA instructions)
7: multiply (2 inside FMA instructions)

The binary function is loading 264 bytes. The binary function is storing 96 bytes.

Arithmetic intensity

Arithmetic intensity is 0.04 FP operations per loaded or stored byte.

General properties

nb instructions	87
loop length	348
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

9.38 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	3.50	3.50	5.50	5.25	8.00	5.25	3.25	3.25	3.25	3.25	12.17	11.83	12.00	2.00	2.00
cycles	3.50	3.50	5.50	5.25	8.00	5.25	3.25	3.25	3.25	3.25	12.17	11.83	12.00	2.00	2.00

Cycles executing div or sqrt instructions

Cycles summary

Front-end	9.38
Overall L1	12.17

Vectorization ratios

INT

all	14%
load	25%
store	33%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	0%
fma	NA (no fma vectorizable/vectorized instructions)
other	0%

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

INT+FP

all	25%
load	25%
store	33%
mul	100%
add-sub	50%
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 call to the function takes 12.17 cycles. At this rate:

22% of peak load performance is reached (21.70 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
16% of peak store performance is reached (7.89 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the function is: 439700

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput
STP X29, X30, [SP, #960]!	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X24, X23, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X22, X21, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X20, X19, [SP, #48]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
ADD X29, SP, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ORR X20, XZR, X3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ORR X21, XZR, X2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ORR X19, XZR, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
BL 43aec0 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE9recomputeERNS1_11ParticleSetE>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR W8, [X19, #1156]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CMP W8, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
B.NE 4397a0 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE11evaluateLogERNS1_11ParticleSetERNS1_14ParticleAttribINS1_10TinyVectorIdLj3EEESaIS9_EEERNS7_IdSaIdEEE+0xa0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR X8, [X19, #256]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDRSW X9, [X19, #1144]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X10, [X21, #24]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
MOVZ W11, #24	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SMADDL X10, W9, W11, X10	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
LDR D0, [X8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDR X8, [X19, #320]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR Q3, [X10]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDR Q2, [X8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDR D1, [X8, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDR X8, [X19, #384]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
FMUL V2.2D, V2.2D, V0.D[0]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	3	0.25
FMUL D1, D1, D0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	3	0.25
FADD V3.2D, V3.2D, V2.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25
FMUL V2.2D, V2.2D, V2.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	3	0.25
STR Q3, [X10]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
LDR D3, [X10, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
FADDP D2, V2.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25
FADD D3, D3, D1	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25
FMADD D1, D1, D1, D2	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	4	0.25
STR D3, [X10, #16]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
LDR D3, [X8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDR X8, [X20, #24]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR D4, [X8, X9,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
FNMSUB D0, D3, D0, D1	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	4	0.25
FADD D0, D0, D4	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25
STR D0, [X8, X9,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
B 4399fc <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE11evaluateLogERNS1_11ParticleSetERNS1_14ParticleAttribINS1_10TinyVectorIdLj3EEESaIS9_EEERNS7_IdSaIdEEE+0x2fc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
CMP W8, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
B.LT 4399fc <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE11evaluateLogERNS1_11ParticleSetERNS1_14ParticleAttribINS1_10TinyVectorIdLj3EEESaIS9_EEERNS7_IdSaIdEEE+0x2fc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR W9, [X19, #1152]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CMP W9, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
B.LT 4399fc <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE11evaluateLogERNS1_11ParticleSetERNS1_14ParticleAttribINS1_10TinyVectorIdLj3EEESaIS9_EEERNS7_IdSaIdEEE+0x2fc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR X17, [X19, #280]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X1, [X19, #216]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X4, [X19, #344]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
ORR X10, XZR, XZR	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
CNTW X16, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
RDVL X18, #3	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
LDR X11, [X19, #256]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X12, [X19, #320]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
RDVL X3, #2	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
LDR X13, [X19, #384]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X14, [X21, #24]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
DUP Z0.D, #0	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	2	0.50
RDVL X6, #1	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
LDR X15, [X20, #24]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
PTRUE P0.D, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
LDRSW X0, [X19, #1144]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
MOVZ W5, #24	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
UBFM X1, X1, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X2, X17, X17,LSL #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
RDVL X17, #6	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
PTRUE P1.B, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
UBFM X4, X4, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
UBFM X2, X2, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
B 43987c <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE11evaluateLogERNS1_11ParticleSetERNS1_14ParticleAttribINS1_10TinyVectorIdLj3EEESaIS9_EEERNS7_IdSaIdEEE+0x17c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0
LDR D0, [X19, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDP X20, X19, [SP, #48]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X22, X21, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X24, X23, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X29, X30, [SP], #64	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
RET	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
HINT #0
HINT #0
HINT #0

Source file and lines	DiracDeterminantRef.cpp:235-255
Module	exec

The function is defined in:

/home/hbollore/qaas/qaas-runs/169-816-7419/intel/miniqmc/build/miniqmc/src/Platforms/CPU/SIMD/inner_product.hpp:81-82,154
/home/hbollore/qaas/qaas-runs/169-816-7419/intel/miniqmc/build/miniqmc/src/Numerics/PETE/OperatorTags.h:63,94
/home/hbollore/qaas/qaas-runs/169-816-7419/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsVector.h:223
/home/hbollore/qaas/qaas-runs/169-816-7419/intel/miniqmc/build/miniqmc/src/QMCWaveFunctions/DiracDeterminantRef.cpp:235-255

Warnings:
Ignoring paths for analysis

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 12.17 to 4.17 cycles (2.92x speedup).

Workaround

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

Vectorization

Details

25% of VPU instructions are used in vector version (process two or more data elements in vector registers):

25% of VPU loads are used in vector version.
33% of VPU stores are used in vector version.
50% of VPU addition or subtraction instructions are used in vector version.
0% 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 function can use their full power.

Workaround

Try another compiler or update/tune your current one
Make array accesses unit-stride:
- If your function 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 12.17 to 0.50 cycles (24.33x speedup).

FMA

Detected 2 FMA (fused multiply-add) operations. Presence of both ADD/SUB and MUL operations.

Workaround

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

The binary function is composed of 14 FP arithmetical operations:

7: addition or subtraction (2 inside FMA instructions)
7: multiply (2 inside FMA instructions)

The binary function is loading 264 bytes. The binary function is storing 96 bytes.

Arithmetic intensity

Arithmetic intensity is 0.04 FP operations per loaded or stored byte.

General properties

nb instructions	87
loop length	348
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

9.38 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	3.50	3.50	5.50	5.25	8.00	5.25	3.25	3.25	3.25	3.25	12.17	11.83	12.00	2.00	2.00
cycles	3.50	3.50	5.50	5.25	8.00	5.25	3.25	3.25	3.25	3.25	12.17	11.83	12.00	2.00	2.00

Cycles executing div or sqrt instructions

Cycles summary

Front-end	9.38
Overall L1	12.17

Vectorization ratios

INT

all	14%
load	25%
store	33%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	0%
fma	NA (no fma vectorizable/vectorized instructions)
other	0%

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

INT+FP

all	25%
load	25%
store	33%
mul	100%
add-sub	50%
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 call to the function takes 12.17 cycles. At this rate:

22% of peak load performance is reached (21.70 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
16% of peak store performance is reached (7.89 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the function is: 439700

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput
STP X29, X30, [SP, #960]!	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X24, X23, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X22, X21, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X20, X19, [SP, #48]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
ADD X29, SP, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ORR X20, XZR, X3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ORR X21, XZR, X2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ORR X19, XZR, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
BL 43aec0 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE9recomputeERNS1_11ParticleSetE>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR W8, [X19, #1156]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CMP W8, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
B.NE 4397a0 <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE11evaluateLogERNS1_11ParticleSetERNS1_14ParticleAttribINS1_10TinyVectorIdLj3EEESaIS9_EEERNS7_IdSaIdEEE+0xa0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR X8, [X19, #256]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDRSW X9, [X19, #1144]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X10, [X21, #24]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
MOVZ W11, #24	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SMADDL X10, W9, W11, X10	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
LDR D0, [X8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDR X8, [X19, #320]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR Q3, [X10]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDR Q2, [X8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDR D1, [X8, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDR X8, [X19, #384]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
FMUL V2.2D, V2.2D, V0.D[0]	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	3	0.25
FMUL D1, D1, D0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	3	0.25
FADD V3.2D, V3.2D, V2.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25
FMUL V2.2D, V2.2D, V2.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	3	0.25
STR Q3, [X10]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
LDR D3, [X10, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
FADDP D2, V2.2D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25
FADD D3, D3, D1	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25
FMADD D1, D1, D1, D2	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	4	0.25
STR D3, [X10, #16]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
LDR D3, [X8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDR X8, [X20, #24]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR D4, [X8, X9,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
FNMSUB D0, D3, D0, D1	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	4	0.25
FADD D0, D0, D4	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25
STR D0, [X8, X9,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
B 4399fc <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE11evaluateLogERNS1_11ParticleSetERNS1_14ParticleAttribINS1_10TinyVectorIdLj3EEESaIS9_EEERNS7_IdSaIdEEE+0x2fc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
CMP W8, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
B.LT 4399fc <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE11evaluateLogERNS1_11ParticleSetERNS1_14ParticleAttribINS1_10TinyVectorIdLj3EEESaIS9_EEERNS7_IdSaIdEEE+0x2fc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR W9, [X19, #1152]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CMP W9, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
B.LT 4399fc <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE11evaluateLogERNS1_11ParticleSetERNS1_14ParticleAttribINS1_10TinyVectorIdLj3EEESaIS9_EEERNS7_IdSaIdEEE+0x2fc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR X17, [X19, #280]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X1, [X19, #216]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X4, [X19, #344]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
ORR X10, XZR, XZR	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
CNTW X16, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
RDVL X18, #3	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
LDR X11, [X19, #256]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X12, [X19, #320]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
RDVL X3, #2	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
LDR X13, [X19, #384]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X14, [X21, #24]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
DUP Z0.D, #0	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	2	0.50
RDVL X6, #1	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
LDR X15, [X20, #24]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
PTRUE P0.D, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
LDRSW X0, [X19, #1144]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
MOVZ W5, #24	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
UBFM X1, X1, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X2, X17, X17,LSL #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
RDVL X17, #6	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
PTRUE P1.B, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
UBFM X4, X4, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
UBFM X2, X2, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
B 43987c <_ZN16miniqmcreference19DiracDeterminantRefIN11qmcplusplus13DelayedUpdateIddEEE11evaluateLogERNS1_11ParticleSetERNS1_14ParticleAttribINS1_10TinyVectorIdLj3EEESaIS9_EEERNS7_IdSaIdEEE+0x17c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0
LDR D0, [X19, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDP X20, X19, [SP, #48]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X22, X21, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X24, X23, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X29, X30, [SP], #64	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
RET	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
HINT #0
HINT #0
HINT #0

Name	Coverage (%)	Time (s)
▼miniqmcreference::DiracDeterminantRef >::evaluateLog(qmcplusplus::ParticleSet&, qmcplusplus::ParticleAttrib, std::allocator > >&, qmcplusplus...–	0.81	0
▼Loop 1045 - inner_product.hpp:81-154 - exec–	0	0
▼Loop 1048 - inner_product.hpp:81-154 - exec–	0	0
○Loop 1047 - inner_product.hpp:81-82 - exec	0.81	0
○Loop 1046 - inner_product.hpp:154-154 - exec	0	0
○Loop 1049 - inner_product.hpp:154-154 - exec	0	0

Report Configuration

Code clean check

Workaround

Vectorization

Details

Workaround

Expensive FP math instructions/calls

FMA

Workaround

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

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

Workaround

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

Arithmetic intensity

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Cycles and memory resources usage

ASM code