OV - exec - Loop 77

	Loop Id: 77	Module: exec	Source: advec_cell_kernel.f90:202-246 [...]	Coverage: 7.69%

0x41e120 ORR	X5, XZR, X8

0x41e124 ORR	X9, XZR, X19

0x41e128 LDR	D21, [X14, X1,LSL #3]    [3]

0x41e12c ORR	X7, XZR, X12

0x41e130 ORR	X6, XZR, X8

0x41e134 FCMPE	D21, #0

0x41e138 B.GT	41e154

0x41e13c CMP	W27, W20

0x41e140 CSEL	W7, W27, W20, #13

0x41e144 SBFM	X6, X7, #0, #31

0x41e148 ORR	X9, XZR, X6

0x41e14c ORR	X7, XZR, X8

0x41e150 ORR	X5, XZR, X12

0x41e154 MADD	X3, X23, X5, X28

0x41e158 FABS	D1, D21

0x41e15c ADD	X2, X11, X1

0x41e160 MADD	X4, X22, X5, X11

0x41e164 ADD	X6, X24, X6

0x41e168 LDR	D2, [X18, X8,LSL #3]    [8]

0x41e16c MADD	X17, X22, X7, X2

0x41e170 ADD	X3, X3, X1

0x41e174 LDR	D31, [X25, X6,LSL #3]    [5]

0x41e178 MADD	X2, X22, X9, X2

0x41e17c ADD	X4, X4, X1

0x41e180 LDR	D29, [X13, X3,LSL #3]    [2]

0x41e184 LDR	D30, [X0, X4,LSL #3]    [6]

0x41e188 LDR	D22, [X0, X17,LSL #3]    [6]

0x41e18c FDIV	D4, D2, D31

0x41e190 LDR	D3, [X0, X2,LSL #3]    [6]

0x41e194 FDIV	D5, D1, D29

0x41e198 FSUB	D6, D22, S30

0x41e19c FSUB	D7, D30, S3

0x41e1a0 FCMPE	D6, #0

0x41e1a4 FABS	D16, D6

0x41e1a8 FMUL	D27, D7, D6

0x41e1ac FABS	D17, D7

0x41e1b0 FMINNM	D18, D16, D17

0x41e1b4 FCSEL	D26, D25, D0, #9

0x41e1b8 FCMPE	D27, #0

0x41e1bc FADD	D28, D5, D0

0x41e1c0 FSUB	D19, D0, S5

0x41e1c4 FSUB	D1, D23, S5

0x41e1c8 FMUL	D2, D4, D28

0x41e1cc B.LS	41e1e8

0x41e1d0 FMUL	D20, D1, D16

0x41e1d4 FMADD	D31, D2, D17, D20

0x41e1d8 FMUL	D29, D31, D24

0x41e1dc FMINNM	D22, D29, D18

0x41e1e0 FMUL	D4, D22, D19

0x41e1e4 FMADD	D30, D26, D4, D30

0x41e1e8 FMUL	D3, D30, D21

0x41e1ec ADD	X6, X30, X1

0x41e1f0 MADD	X5, X21, X5, X6

0x41e1f4 MADD	X7, X21, X7, X6

0x41e1f8 MADD	X9, X21, X9, X6

0x41e1fc STR	D3, [X15, X1,LSL #3]    [4]

0x41e200 FABS	D21, D3

0x41e204 LDR	D5, [X10, X5,LSL #3]    [1]

0x41e208 LDR	D6, [X10, X7,LSL #3]    [1]

0x41e20c LDR	D7, [X10, X9,LSL #3]    [1]

0x41e210 FMUL	D16, D5, D3

0x41e214 FSUB	D17, D6, S5

0x41e218 FSUB	D27, D5, S7

0x41e21c FABS	D26, D17

0x41e220 FCMPE	D17, #0

0x41e224 FMUL	D28, D27, D17

0x41e228 FABS	D19, D27

0x41e22c FMINNM	D18, D26, D19

0x41e230 FMUL	D1, D1, D26

0x41e234 FCSEL	D20, D25, D0, #9

0x41e238 FCMPE	D28, #0

0x41e23c FMADD	D2, D2, D19, D1

0x41e240 B.LS	41ec10

0x41e244 LDR	D31, [X0, X4,LSL #3]    [6]

0x41e248 FMUL	D4, D2, D24

0x41e24c FMINNM	D30, D4, D18

0x41e250 LDR	D29, [X13, X3,LSL #3]    [2]

0x41e254 FMUL	D22, D31, D29

0x41e258 FDIV	D21, D21, D22

0x41e25c FSUB	D6, D0, S21

0x41e260 FMUL	D7, D6, D30

0x41e264 FMADD	D5, D20, D7, D5

0x41e268 FMUL	D3, D5, D3

0x41e26c STR	D3, [X16, X1,LSL #3]    [7]

0x41e270 ADD	X1, X1, #1

0x41e274 CMP	W26, W1

0x41e278 B.GE	41e120

0x41ec10 STR	D16, [X16, X1,LSL #3]    [7]

0x41ec14 ADD	X1, X1, #1

0x41ec18 CMP	W26, W1

0x41ec1c B.GE	41e120

/home/eoseret/qaas/qaas_runs/178-231-1255/intel/CloverLeaf1.3-FC/build/CloverLeaf1.3-FC/CloverLeaf_ref/kernels/advec_cell_kernel.f90: 202 - 246

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

202:         DO j=x_min,x_max

203:

204:           IF(vol_flux_y(j,k).GT.0.0)THEN

[...]

210:             upwind   =MIN(k+1,y_max+2)

[...]

216:           sigmat=ABS(vol_flux_y(j,k))/pre_vol(j,donor)

217:           sigma3=(1.0_8+sigmat)*(vertexdy(k)/vertexdy(dif))

218:           sigma4=2.0_8-sigmat

219:

220:           sigma=sigmat

221:           sigmav=sigmat

222:

223:           diffuw=density1(j,donor)-density1(j,upwind)

224:           diffdw=density1(j,downwind)-density1(j,donor)

225:           wind=1.0_8

226:           IF(diffdw.LE.0.0) wind=-1.0_8

227:           IF(diffuw*diffdw.GT.0.0)THEN

228:             limiter=(1.0_8-sigmav)*wind*MIN(ABS(diffuw),ABS(diffdw)&

229:               ,one_by_six*(sigma3*ABS(diffuw)+sigma4*ABS(diffdw)))

230:           ELSE

231:             limiter=0.0

232:           ENDIF

233:           mass_flux_y(j,k)=vol_flux_y(j,k)*(density1(j,donor)+limiter)

234:

235:           sigmam=ABS(mass_flux_y(j,k))/(density1(j,donor)*pre_vol(j,donor))

236:           diffuw=energy1(j,donor)-energy1(j,upwind)

237:           diffdw=energy1(j,downwind)-energy1(j,donor)

238:           wind=1.0_8

239:           IF(diffdw.LE.0.0) wind=-1.0_8

240:           IF(diffuw*diffdw.GT.0.0)THEN

241:             limiter=(1.0_8-sigmam)*wind*MIN(ABS(diffuw),ABS(diffdw)&

242:               ,one_by_six*(sigma3*ABS(diffuw)+sigma4*ABS(diffdw)))

243:           ELSE

244:             limiter=0.0

245:           ENDIF

246:           ener_flux(j,k)=mass_flux_y(j,k)*(energy1(j,donor)+limiter)

Coverage (%)	Name	Source Location	Module
►50.57+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:88	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec
►49.43+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:57	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec

Coverage (%)	Name	Source Location	Module
►49.98+	omp_fulfill_event		libgomp.so.1.0.0
○	start_thread		libc.so.6
○	thread_start		libc.so.6
►25.30+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:88	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec
►24.72+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:57	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec

Coverage (%)	Name	Source Location	Module
►74.89+	omp_fulfill_event		libgomp.so.1.0.0
○	start_thread		libc.so.6
○	thread_start		libc.so.6
►12.71+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:88	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec
►12.40+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:57	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec

Coverage (%)	Name	Source Location	Module
►87.51+	omp_fulfill_event		libgomp.so.1.0.0
○	start_thread		libc.so.6
○	thread_start		libc.so.6
►6.31+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:88	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec
►6.18+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:57	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec

Coverage (%)	Name	Source Location	Module
►93.76+	omp_fulfill_event		libgomp.so.1.0.0
○	start_thread		libc.so.6
○	thread_start		libc.so.6
►3.15+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:88	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec
►3.08+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:57	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec

Coverage (%)	Name	Source Location	Module
►95.77+	omp_fulfill_event		libgomp.so.1.0.0
○	start_thread		libc.so.6
○	thread_start		libc.so.6
►2.15+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:88	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec
►2.08+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:57	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec

Coverage (%)	Name	Source Location	Module
►96.81+	omp_fulfill_event		libgomp.so.1.0.0
○	start_thread		libc.so.6
○	thread_start		libc.so.6
►1.61+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:88	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec
►1.58+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:57	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec

Coverage (%)	Name	Source Location	Module
►97.45+	omp_fulfill_event		libgomp.so.1.0.0
○	start_thread		libc.so.6
○	thread_start		libc.so.6
►1.29+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:88	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec
►1.26+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:57	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec

Coverage (%)	Name	Source Location	Module
►97.88+	omp_fulfill_event		libgomp.so.1.0.0
○	start_thread		libc.so.6
○	thread_start		libc.so.6
►1.07+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:88	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec
►1.04+	GOMP_parallel		libgomp.so.1.0.0
○	advec_cell_kernel	advec_cell_kernel.f90:270	exec
○	advec_cell_driver	advec_cell_driver.f90:84	exec
○	advection	advection.f90:57	exec
○	hydro	hydro.f90:64	exec
○	main	clover_leaf.f90:41	exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start	clover_leaf.f90:41	exec

Coverage (%)	Name	Module
►98.17+	omp_fulfill_event	libgomp.so.1.0.0
○	start_thread	libc.so.6
○	thread_start	libc.so.6

Coverage (%)	Name	Module
►98.41+	omp_fulfill_event	libgomp.so.1.0.0
○	start_thread	libc.so.6
○	thread_start	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	2.32
CQA speedup if FP arith vectorized	1.78
CQA speedup if fully vectorized	4.00 - 2.29
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.05
Bottlenecks	P6, P7, P8, P9,
Function	__advec_cell_kernel_module_MOD_advec_cell_kernel._omp_fn.0
Source	advec_cell_kernel.f90:202-204,advec_cell_kernel.f90:210-210,advec_cell_kernel.f90:216-246
Source loop unroll info	not unrolled or unrolled with no peel/tail loop
Source loop unroll confidence level	max
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	12.00
CQA cycles if no scalar integer	5.17
CQA cycles if FP arith vectorized	6.75
CQA cycles if fully vectorized	3.00 - 5.24
Front-end cycles	11.38
P0 cycles	2.50
P1 cycles	2.50
P2 cycles	6.50
P3 cycles	6.25
P4 cycles	7.00
P5 cycles	6.25
P6 cycles	12.00
P7 cycles	12.00
P8 cycles	12.00
P9 cycles	12.00
P10 cycles	5.17
P11 cycles	4.83
P12 cycles	5.00
P13 cycles	0.00
P14 cycles	0.00
DIV/SQRT cycles	5.25 - 10.49
Inter-iter dependencies cycles	NA
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	91.00
Nb uops	91.00
Nb loads	NA
Nb stores	3.00
Nb stack references	0.00
FLOP/cycle	2.67
Nb FLOP add-sub	8.00
Nb FLOP mul	13.00
Nb FLOP fma	4.00
Nb FLOP div	3.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	0.00
Bytes prefetched	0.00
Bytes loaded	0.00
Bytes stored	0.00
Stride 0	NA
Stride 1	NA
Stride n	NA
Stride unknown	NA
Stride indirect	NA
Vectorization ratio all	0.00
Vectorization ratio load	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	0.00
Vector-efficiency ratio all	24.60
Vector-efficiency ratio load	25.00
Vector-efficiency ratio store	25.00
Vector-efficiency ratio mul	25.00
Vector-efficiency ratio add_sub	25.00
Vector-efficiency ratio fma	25.00
Vector-efficiency ratio div_sqrt	25.00
Vector-efficiency ratio other	23.68

Metric	Value
CQA speedup if no scalar integer	2.32
CQA speedup if FP arith vectorized	1.78
CQA speedup if fully vectorized	4.00 - 2.29
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.05
Bottlenecks	P6, P7, P8, P9,
Function	__advec_cell_kernel_module_MOD_advec_cell_kernel._omp_fn.0
Source	advec_cell_kernel.f90:202-204,advec_cell_kernel.f90:210-210,advec_cell_kernel.f90:216-246
Source loop unroll info	not unrolled or unrolled with no peel/tail loop
Source loop unroll confidence level	max
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	12.00
CQA cycles if no scalar integer	5.17
CQA cycles if FP arith vectorized	6.75
CQA cycles if fully vectorized	3.00 - 5.24
Front-end cycles	11.38
P0 cycles	2.50
P1 cycles	2.50
P2 cycles	6.50
P3 cycles	6.25
P4 cycles	7.00
P5 cycles	6.25
P6 cycles	12.00
P7 cycles	12.00
P8 cycles	12.00
P9 cycles	12.00
P10 cycles	5.17
P11 cycles	4.83
P12 cycles	5.00
P13 cycles	0.00
P14 cycles	0.00
DIV/SQRT cycles	5.25 - 10.49
Inter-iter dependencies cycles	NA
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	91.00
Nb uops	91.00
Nb loads	NA
Nb stores	3.00
Nb stack references	0.00
FLOP/cycle	2.67
Nb FLOP add-sub	8.00
Nb FLOP mul	13.00
Nb FLOP fma	4.00
Nb FLOP div	3.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	0.00
Bytes prefetched	0.00
Bytes loaded	0.00
Bytes stored	0.00
Stride 0	NA
Stride 1	NA
Stride n	NA
Stride unknown	NA
Stride indirect	NA
Vectorization ratio all	0.00
Vectorization ratio load	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	0.00
Vector-efficiency ratio all	24.60
Vector-efficiency ratio load	25.00
Vector-efficiency ratio store	25.00
Vector-efficiency ratio mul	25.00
Vector-efficiency ratio add_sub	25.00
Vector-efficiency ratio fma	25.00
Vector-efficiency ratio div_sqrt	25.00
Vector-efficiency ratio other	23.68

Path /

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

Function	__advec_cell_kernel_module_MOD_advec_cell_kernel._omp_fn.0
Source file and lines	advec_cell_kernel.f90:202-246
Module	exec

The loop is defined in /home/eoseret/qaas/qaas_runs/178-231-1255/intel/CloverLeaf1.3-FC/build/CloverLeaf1.3-FC/CloverLeaf_ref/kernels/advec_cell_kernel.f90:202-204,210,216-246.

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

Ignoring paths for analysis
Too many paths. Rerun with max-paths=8
RecMII not computed since number of paths is unknown or > max_paths
Streams not analyzed since number of paths is unknown or > max_paths

Try to simplify control and/or increase the maximum number of paths per function/loop through the 'max-paths-nb' option.

This loop has 8 execution paths.

The presence of multiple execution paths is typically the main/first bottleneck.
Try to simplify control inside loop: ideally, try to remove all conditional expressions, for example by (if applicable):

hoisting them (moving them outside the loop)
turning them into conditional moves, MIN or MAX

Ex: if (x<0) x=0 => x = max(0,x) (Fortran instrinsic procedure)

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 12.00 to 5.17 cycles (2.32x speedup).

Workaround

Try to reorganize arrays of structures to structures of arrays
Consider to permute loops (see vectorization gain report)
To reference allocatable arrays, use "allocatable" instead of "pointer" pointers or qualify them with the "contiguous" attribute (Fortran 2008)
For structures, limit to one indirection. For example, use a_b%c instead of a%b%c with a_b set to a%b before this loop

Vectorization

Your loop is not vectorized. Only 24% of vector register length is used (average across all VPU instructions). By vectorizing your loop, you can lower the cost of an iteration from 12.00 to 5.24 cycles (2.29x speedup).

Details

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

Workaround

Try another compiler or update/tune your current one
Remove inter-iterations dependences from your loop and make it unit-stride:
- If your arrays have 2 or more dimensions, check whether elements are accessed contiguously and, otherwise, try to permute loops accordingly: Fortran storage order is column-major: do i do j a(i,j) = b(i,j) (slow, non stride 1) => do i do j a(j,i) = b(i,j) (fast, stride 1)
- If your loop streams arrays of structures (AoS), try to use structures of arrays instead (SoA): do i a(i)%x = b(i)%x (slow, non stride 1) => do 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 12.00 to 11.25 cycles (1.07x speedup).

FMA

Detected 4 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 loop (in the source code) and the binary loop

The binary loop is composed of 32 FP arithmetical operations:

12: addition or subtraction (4 inside FMA instructions)
17: multiply (4 inside FMA instructions)
3: divide

The binary loop does not load or store any data.

General properties

nb instructions	91
nb uops	91
loop length	364
used w registers	5
used x registers	28
used b registers	0
used h registers	0
used s registers	5
used d registers	24
used q registers	0
used v registers	0
used z registers	0
nb stack references	0
ADD-SUB / MUL ratio	0.62

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	11.38 cycles
front end	11.38 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	2.50	2.50	6.50	6.25	7.00	6.25	12.00	12.00	12.00	12.00	5.17	4.83	5.00	0.00	0.00
cycles	2.50	2.50	6.50	6.25	7.00	6.25	12.00	12.00	12.00	12.00	5.17	4.83	5.00	0.00	0.00

Execution ports to units layout:

P0: BRU
P1: BRU
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

5.25-10.49

Cycles summary

Front-end	11.38
Dispatch	12.00
DIV/SQRT	5.25-10.49
Overall L1	12.00

Vectorization ratios

INT

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

all	0%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	0%
add-sub	0%
fma	0%
div/sqrt	0%
other	0%

INT+FP

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

Vector efficiency ratios

INT

all	23%
load	25%
store	25%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	NA (no add-sub vectorizable/vectorized instructions)
fma	NA (no fma vectorizable/vectorized instructions)
other	12%

all	25%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	25%
add-sub	25%
fma	25%
div/sqrt	25%
other	25%

INT+FP

all	24%
load	25%
store	25%
mul	25%
add-sub	25%
fma	25%
div/sqrt	25%
other	23%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 12.00 cycles.

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the loop is: 41e120

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	Latency	Recip. throughput	Vectorization
ORR X5, XZR, X8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X9, XZR, X19	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
LDR D21, [X14, X1,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
ORR X7, XZR, X12	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X6, XZR, X8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
FCMPE D21, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
B.GT 41e154 <__advec_cell_kernel_module_MOD_advec_cell_kernel._omp_fn.0+0x394>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP W27, W20	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
CSEL W7, W27, W20, #13	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
SBFM X6, X7, #0, #31	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X9, XZR, X6	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X7, XZR, X8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X5, XZR, X12	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
MADD X3, X23, X5, X28	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
FABS D1, D21	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
ADD X2, X11, X1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
MADD X4, X22, X5, X11	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X6, X24, X6	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
LDR D2, [X18, X8,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
MADD X17, X22, X7, X2	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X3, X3, X1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
LDR D31, [X25, X6,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
MADD X2, X22, X9, X2	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X4, X4, X1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
LDR D29, [X13, X3,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D30, [X0, X4,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D22, [X0, X17,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FDIV D4, D2, D31	1	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	1.75-3.50	scal (25.0%)
LDR D3, [X0, X2,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FDIV D5, D1, D29	1	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	1.75-3.50	scal (25.0%)
FSUB D6, D22, S30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D7, D30, S3	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FCMPE D6, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
FABS D16, D6	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D27, D7, D6	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FABS D17, D7	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMINNM D18, D16, D17	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FCSEL D26, D25, D0, #9	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	2	0.50	scal (25.0%)
FCMPE D27, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
FADD D28, D5, D0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D19, D0, S5	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D1, D23, S5	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D2, D4, D28	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
B.LS 41e1e8 <__advec_cell_kernel_module_MOD_advec_cell_kernel._omp_fn.0+0x428>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
FMUL D20, D1, D16	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMADD D31, D2, D17, D20	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	scal (25.0%)
FMUL D29, D31, D24	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMINNM D22, D29, D18	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D4, D22, D19	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMADD D30, D26, D4, D30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	scal (25.0%)
FMUL D3, D30, D21	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
ADD X6, X30, X1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
MADD X5, X21, X5, X6	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X7, X21, X7, X6	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X9, X21, X9, X6	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
STR D3, [X15, X1,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (25.0%)
FABS D21, D3	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
LDR D5, [X10, X5,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D6, [X10, X7,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D7, [X10, X9,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FMUL D16, D5, D3	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FSUB D17, D6, S5	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D27, D5, S7	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FABS D26, D17	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FCMPE D17, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
FMUL D28, D27, D17	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FABS D19, D27	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMINNM D18, D26, D19	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D1, D1, D26	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FCSEL D20, D25, D0, #9	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	2	0.50	scal (25.0%)
FCMPE D28, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
FMADD D2, D2, D19, D1	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	scal (25.0%)
B.LS 41ec10 <__advec_cell_kernel_module_MOD_advec_cell_kernel._omp_fn.0+0xe50>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR D31, [X0, X4,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FMUL D4, D2, D24	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMINNM D30, D4, D18	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
LDR D29, [X13, X3,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FMUL D22, D31, D29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FDIV D21, D21, D22	1	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	1.75-3.50	scal (25.0%)
FSUB D6, D0, S21	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D7, D6, D30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMADD D5, D20, D7, D5	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	scal (25.0%)
FMUL D3, D5, D3	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
STR D3, [X16, X1,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (25.0%)
ADD X1, X1, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
CMP W26, W1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
B.GE 41e120 <__advec_cell_kernel_module_MOD_advec_cell_kernel._omp_fn.0+0x360>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
STR D16, [X16, X1,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (25.0%)
ADD X1, X1, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
CMP W26, W1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	N/A
B.GE 41e120 <__advec_cell_kernel_module_MOD_advec_cell_kernel._omp_fn.0+0x360>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A

Function	__advec_cell_kernel_module_MOD_advec_cell_kernel._omp_fn.0
Source file and lines	advec_cell_kernel.f90:202-246
Module	exec

Ignoring paths for analysis
Too many paths. Rerun with max-paths=8
RecMII not computed since number of paths is unknown or > max_paths
Streams not analyzed since number of paths is unknown or > max_paths

hoisting them (moving them outside the loop)
turning them into conditional moves, MIN or MAX

Ex: if (x<0) x=0 => x = max(0,x) (Fortran instrinsic procedure)

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 12.00 to 5.17 cycles (2.32x speedup).

Workaround

Try to reorganize arrays of structures to structures of arrays
Consider to permute loops (see vectorization gain report)
To reference allocatable arrays, use "allocatable" instead of "pointer" pointers or qualify them with the "contiguous" attribute (Fortran 2008)
For structures, limit to one indirection. For example, use a_b%c instead of a%b%c with a_b set to a%b before this loop

Vectorization

Details

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

Workaround

Try another compiler or update/tune your current one
Remove inter-iterations dependences from your loop and make it unit-stride:
- If your arrays have 2 or more dimensions, check whether elements are accessed contiguously and, otherwise, try to permute loops accordingly: Fortran storage order is column-major: do i do j a(i,j) = b(i,j) (slow, non stride 1) => do i do j a(j,i) = b(i,j) (fast, stride 1)
- If your loop streams arrays of structures (AoS), try to use structures of arrays instead (SoA): do i a(i)%x = b(i)%x (slow, non stride 1) => do 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 12.00 to 11.25 cycles (1.07x speedup).

FMA

Detected 4 FMA (fused multiply-add) operations. Presence of both ADD/SUB and MUL operations.

Workaround

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

The binary loop is composed of 32 FP arithmetical operations:

12: addition or subtraction (4 inside FMA instructions)
17: multiply (4 inside FMA instructions)
3: divide

The binary loop does not load or store any data.

General properties

nb instructions	91
nb uops	91
loop length	364
used w registers	5
used x registers	28
used b registers	0
used h registers	0
used s registers	5
used d registers	24
used q registers	0
used v registers	0
used z registers	0
nb stack references	0
ADD-SUB / MUL ratio	0.62

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	11.38 cycles
front end	11.38 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	2.50	2.50	6.50	6.25	7.00	6.25	12.00	12.00	12.00	12.00	5.17	4.83	5.00	0.00	0.00
cycles	2.50	2.50	6.50	6.25	7.00	6.25	12.00	12.00	12.00	12.00	5.17	4.83	5.00	0.00	0.00

Execution ports to units layout:

P0: BRU
P1: BRU
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

5.25-10.49

Cycles summary

Front-end	11.38
Dispatch	12.00
DIV/SQRT	5.25-10.49
Overall L1	12.00

Vectorization ratios

INT

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

all	0%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	0%
add-sub	0%
fma	0%
div/sqrt	0%
other	0%

INT+FP

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

Vector efficiency ratios

INT

all	23%
load	25%
store	25%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	NA (no add-sub vectorizable/vectorized instructions)
fma	NA (no fma vectorizable/vectorized instructions)
other	12%

all	25%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	25%
add-sub	25%
fma	25%
div/sqrt	25%
other	25%

INT+FP

all	24%
load	25%
store	25%
mul	25%
add-sub	25%
fma	25%
div/sqrt	25%
other	23%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 12.00 cycles.

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the loop is: 41e120

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	Latency	Recip. throughput	Vectorization
ORR X5, XZR, X8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X9, XZR, X19	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
LDR D21, [X14, X1,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
ORR X7, XZR, X12	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X6, XZR, X8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
FCMPE D21, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
B.GT 41e154 <__advec_cell_kernel_module_MOD_advec_cell_kernel._omp_fn.0+0x394>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP W27, W20	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
CSEL W7, W27, W20, #13	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
SBFM X6, X7, #0, #31	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X9, XZR, X6	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X7, XZR, X8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X5, XZR, X12	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
MADD X3, X23, X5, X28	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
FABS D1, D21	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
ADD X2, X11, X1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
MADD X4, X22, X5, X11	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X6, X24, X6	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
LDR D2, [X18, X8,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
MADD X17, X22, X7, X2	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X3, X3, X1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
LDR D31, [X25, X6,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
MADD X2, X22, X9, X2	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
ADD X4, X4, X1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
LDR D29, [X13, X3,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D30, [X0, X4,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D22, [X0, X17,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FDIV D4, D2, D31	1	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	1.75-3.50	scal (25.0%)
LDR D3, [X0, X2,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FDIV D5, D1, D29	1	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	1.75-3.50	scal (25.0%)
FSUB D6, D22, S30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D7, D30, S3	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FCMPE D6, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
FABS D16, D6	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D27, D7, D6	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FABS D17, D7	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMINNM D18, D16, D17	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FCSEL D26, D25, D0, #9	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	2	0.50	scal (25.0%)
FCMPE D27, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
FADD D28, D5, D0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D19, D0, S5	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D1, D23, S5	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D2, D4, D28	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
B.LS 41e1e8 <__advec_cell_kernel_module_MOD_advec_cell_kernel._omp_fn.0+0x428>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
FMUL D20, D1, D16	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMADD D31, D2, D17, D20	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	scal (25.0%)
FMUL D29, D31, D24	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMINNM D22, D29, D18	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D4, D22, D19	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMADD D30, D26, D4, D30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	scal (25.0%)
FMUL D3, D30, D21	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
ADD X6, X30, X1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
MADD X5, X21, X5, X6	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X7, X21, X7, X6	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X9, X21, X9, X6	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
STR D3, [X15, X1,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (25.0%)
FABS D21, D3	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
LDR D5, [X10, X5,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D6, [X10, X7,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D7, [X10, X9,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FMUL D16, D5, D3	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FSUB D17, D6, S5	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D27, D5, S7	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FABS D26, D17	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FCMPE D17, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
FMUL D28, D27, D17	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FABS D19, D27	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMINNM D18, D26, D19	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D1, D1, D26	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FCSEL D20, D25, D0, #9	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	2	0.50	scal (25.0%)
FCMPE D28, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
FMADD D2, D2, D19, D1	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	scal (25.0%)
B.LS 41ec10 <__advec_cell_kernel_module_MOD_advec_cell_kernel._omp_fn.0+0xe50>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR D31, [X0, X4,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FMUL D4, D2, D24	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMINNM D30, D4, D18	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
LDR D29, [X13, X3,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FMUL D22, D31, D29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FDIV D21, D21, D22	1	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	1.75-3.50	scal (25.0%)
FSUB D6, D0, S21	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D7, D6, D30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMADD D5, D20, D7, D5	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	scal (25.0%)
FMUL D3, D5, D3	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
STR D3, [X16, X1,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (25.0%)
ADD X1, X1, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
CMP W26, W1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
B.GE 41e120 <__advec_cell_kernel_module_MOD_advec_cell_kernel._omp_fn.0+0x360>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
STR D16, [X16, X1,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (25.0%)
ADD X1, X1, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
CMP W26, W1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	N/A
B.GE 41e120 <__advec_cell_kernel_module_MOD_advec_cell_kernel._omp_fn.0+0x360>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A

▶Scalability Runs Description

Run 1x1	Number processes: 1Number nodes: NARun Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to core --map-by package:PE=64 --rank-by fill --report-bindings Dataset: Run Directory: /home/eoseret/qaas/qaas_runs/178-231-1255/intel/CloverLeaf1.3-FC/run/oneview_runs/multicore/gcc_5/oneview_run_1782340583OMP_PROC_BIND: spreadOMP_DISPLAY_AFFINITY: TRUEOMP_AFFINITY_FORMAT: 'OMP: pid %P tid %i thread %n bound to OS proc set {%A}'OMP_DISPLAY_ENV: TRUEOMP_NUM_THREADS: 1OMP_PLACES: threads
Run 1x2	Number processes: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to core --map-by package:PE=64 --rank-by fill --report-bindings Dataset: Run Directory: /home/eoseret/qaas/qaas_runs/178-231-1255/intel/CloverLeaf1.3-FC/run/oneview_runs/multicore/gcc_5/oneview_run_1782340583OMP_NUM_THREADS: 2OMP_PROC_BIND: spreadOMP_DISPLAY_AFFINITY: TRUEOMP_AFFINITY_FORMAT: 'OMP: pid %P tid %i thread %n bound to OS proc set {%A}'OMP_DISPLAY_ENV: TRUEOMP_PLACES: threads
Run 1x4	Number processes: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to core --map-by package:PE=64 --rank-by fill --report-bindings Dataset: Run Directory: /home/eoseret/qaas/qaas_runs/178-231-1255/intel/CloverLeaf1.3-FC/run/oneview_runs/multicore/gcc_5/oneview_run_1782340583OMP_NUM_THREADS: 4OMP_PROC_BIND: spreadOMP_DISPLAY_AFFINITY: TRUEOMP_AFFINITY_FORMAT: 'OMP: pid %P tid %i thread %n bound to OS proc set {%A}'OMP_DISPLAY_ENV: TRUEOMP_PLACES: threads
Run 1x8	Number processes: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to core --map-by package:PE=64 --rank-by fill --report-bindings Dataset: Run Directory: /home/eoseret/qaas/qaas_runs/178-231-1255/intel/CloverLeaf1.3-FC/run/oneview_runs/multicore/gcc_5/oneview_run_1782340583OMP_NUM_THREADS: 8OMP_PROC_BIND: spreadOMP_DISPLAY_AFFINITY: TRUEOMP_AFFINITY_FORMAT: 'OMP: pid %P tid %i thread %n bound to OS proc set {%A}'OMP_DISPLAY_ENV: TRUEOMP_PLACES: threads
Run 1x16	Number processes: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to core --map-by package:PE=64 --rank-by fill --report-bindings Dataset: Run Directory: /home/eoseret/qaas/qaas_runs/178-231-1255/intel/CloverLeaf1.3-FC/run/oneview_runs/multicore/gcc_5/oneview_run_1782340583OMP_NUM_THREADS: 16OMP_PROC_BIND: spreadOMP_DISPLAY_AFFINITY: TRUEOMP_AFFINITY_FORMAT: 'OMP: pid %P tid %i thread %n bound to OS proc set {%A}'OMP_DISPLAY_ENV: TRUEOMP_PLACES: threads
Run 1x24	Number processes: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to core --map-by package:PE=64 --rank-by fill --report-bindings Dataset: Run Directory: /home/eoseret/qaas/qaas_runs/178-231-1255/intel/CloverLeaf1.3-FC/run/oneview_runs/multicore/gcc_5/oneview_run_1782340583OMP_NUM_THREADS: 24OMP_PROC_BIND: spreadOMP_DISPLAY_AFFINITY: TRUEOMP_AFFINITY_FORMAT: 'OMP: pid %P tid %i thread %n bound to OS proc set {%A}'OMP_DISPLAY_ENV: TRUEOMP_PLACES: threads
Run 1x32	Number processes: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to core --map-by package:PE=64 --rank-by fill --report-bindings Dataset: Run Directory: /home/eoseret/qaas/qaas_runs/178-231-1255/intel/CloverLeaf1.3-FC/run/oneview_runs/multicore/gcc_5/oneview_run_1782340583OMP_NUM_THREADS: 32OMP_PROC_BIND: spreadOMP_DISPLAY_AFFINITY: TRUEOMP_AFFINITY_FORMAT: 'OMP: pid %P tid %i thread %n bound to OS proc set {%A}'OMP_DISPLAY_ENV: TRUEOMP_PLACES: threads
Run 1x40	Number processes: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to core --map-by package:PE=64 --rank-by fill --report-bindings Dataset: Run Directory: /home/eoseret/qaas/qaas_runs/178-231-1255/intel/CloverLeaf1.3-FC/run/oneview_runs/multicore/gcc_5/oneview_run_1782340583OMP_NUM_THREADS: 40OMP_PROC_BIND: spreadOMP_DISPLAY_AFFINITY: TRUEOMP_AFFINITY_FORMAT: 'OMP: pid %P tid %i thread %n bound to OS proc set {%A}'OMP_DISPLAY_ENV: TRUEOMP_PLACES: threads
Run 1x48	Number processes: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to core --map-by package:PE=64 --rank-by fill --report-bindings Dataset: Run Directory: /home/eoseret/qaas/qaas_runs/178-231-1255/intel/CloverLeaf1.3-FC/run/oneview_runs/multicore/gcc_5/oneview_run_1782340583OMP_NUM_THREADS: 48OMP_PROC_BIND: spreadOMP_DISPLAY_AFFINITY: TRUEOMP_AFFINITY_FORMAT: 'OMP: pid %P tid %i thread %n bound to OS proc set {%A}'OMP_DISPLAY_ENV: TRUEOMP_PLACES: threads
Run 1x56	Number processes: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to core --map-by package:PE=64 --rank-by fill --report-bindings Dataset: Run Directory: /home/eoseret/qaas/qaas_runs/178-231-1255/intel/CloverLeaf1.3-FC/run/oneview_runs/multicore/gcc_5/oneview_run_1782340583OMP_NUM_THREADS: 56OMP_PROC_BIND: spreadOMP_DISPLAY_AFFINITY: TRUEOMP_AFFINITY_FORMAT: 'OMP: pid %P tid %i thread %n bound to OS proc set {%A}'OMP_DISPLAY_ENV: TRUEOMP_PLACES: threads
Run 1x64	Number processes: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to core --map-by package:PE=64 --rank-by fill --report-bindings Dataset: Run Directory: /home/eoseret/qaas/qaas_runs/178-231-1255/intel/CloverLeaf1.3-FC/run/oneview_runs/multicore/gcc_5/oneview_run_1782340583OMP_NUM_THREADS: 64OMP_PROC_BIND: spreadOMP_DISPLAY_AFFINITY: TRUEOMP_AFFINITY_FORMAT: 'OMP: pid %P tid %i thread %n bound to OS proc set {%A}'OMP_DISPLAY_ENV: TRUEOMP_PLACES: threads

(1x1) Efficiency	(1x1) Potential Speed-Up (%)	(1x2) Efficiency	(1x2) Potential Speed-Up (%)	(1x4) Efficiency	(1x4) Potential Speed-Up (%)	(1x8) Efficiency	(1x8) Potential Speed-Up (%)	(1x16) Efficiency	(1x16) Potential Speed-Up (%)	(1x24) Efficiency	(1x24) Potential Speed-Up (%)	(1x32) Efficiency	(1x32) Potential Speed-Up (%)	(1x40) Efficiency	(1x40) Potential Speed-Up (%)	(1x48) Efficiency	(1x48) Potential Speed-Up (%)	(1x56) Efficiency	(1x56) Potential Speed-Up (%)	(1x64) Efficiency	(1x64) Potential Speed-Up (%)
1	0	0.99	0.04	0.99	0.1	0.98	0.13	0.96	0.18	0.88	0.4	0.68	1.13	0.56	1.55	0.47	1.92	0.4	2.22	0.34	2.46

Speed-Up and Efficiency

▶Detailed Speed-Up and Efficiency

Run	Number of threads	Efficiency (ideal is 1)	Speedup	Ideal Speedup	Time (s)	Coverage (%)
1x1	1	1	1	1	107.59998321533	7.692934513092
1x2	2	0.99	1.99	2	53.980003356934	7.6538209915161
1x4	4	0.99	3.95	4	27.11499786377	7.4897413253784
1x8	8	0.98	7.84	8	13.600001335144	6.520866394043
1x16	16	0.96	15.32	16	7.0499992370605	4.353178024292
1x24	24	0.88	21.14	24	5.2049999237061	3.34348320961
1x32	32	0.68	21.73	32	4.9850006103516	3.515837430954
1x40	40	0.56	22.59	40	4.8350005149841	3.562691450119
1x48	48	0.47	22.6	48	4.7950000762939	3.6303884983063
1x56	56	0.4	22.31	56	4.9000005722046	3.6921367645264
1x64	64	0.34	21.93	64	4.9650001525879	3.749386548996

Report Configuration

Code clean check

Workaround

Vectorization

Details

Workaround

Expensive FP math instructions/calls

FMA

Workaround

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

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code

Code clean check

Workaround

Vectorization

Details

Workaround

Expensive FP math instructions/calls

FMA

Workaround

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

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code

▶Scalability Runs Description

▼Columns Filter

Speed-Up and Efficiency

▶Detailed Speed-Up and Efficiency