OV - exec - Loop 228

	Loop Id: 228	Module: exec	Source: advec_mom.cpp:88-88	Coverage: 1.00%

0x4333d7 VMOVUPD	(%R11,%RAX,1),%ZMM8    [3]

0x4333de VMOVUPD	(%R9,%RAX,1),%ZMM10    [1]

0x4333e5 VADDPD	(%R10,%RAX,1),%ZMM8,%ZMM9    [2]

0x4333ec VADDPD	(%R8,%RAX,1),%ZMM10,%ZMM11    [5]

0x4333f3 VADDPD	%ZMM11,%ZMM9,%ZMM12

0x4333f9 VMULPD	%ZMM1,%ZMM12,%ZMM13

0x4333ff VMOVUPD	%ZMM13,(%RDI,%RAX,1)    [4]

0x433406 VMOVUPD	0x40(%R11,%RAX,1),%ZMM14    [3]

0x43340e VMOVUPD	0x40(%R9,%RAX,1),%ZMM0    [1]

0x433416 VADDPD	0x40(%R10,%RAX,1),%ZMM14,%ZMM15    [2]

0x43341e VADDPD	0x40(%R8,%RAX,1),%ZMM0,%ZMM5    [5]

0x433426 VADDPD	%ZMM5,%ZMM15,%ZMM2

0x43342c VMULPD	%ZMM1,%ZMM2,%ZMM6

0x433432 VMOVUPD	%ZMM6,0x40(%RDI,%RAX,1)    [4]

0x43343a VMOVUPD	0x80(%R11,%RAX,1),%ZMM7    [3]

0x433442 VMOVUPD	0x80(%R9,%RAX,1),%ZMM9    [1]

0x43344a VADDPD	0x80(%R10,%RAX,1),%ZMM7,%ZMM8    [2]

0x433452 VADDPD	0x80(%R8,%RAX,1),%ZMM9,%ZMM10    [5]

0x43345a VADDPD	%ZMM10,%ZMM8,%ZMM11

0x433460 VMULPD	%ZMM1,%ZMM11,%ZMM12

0x433466 VMOVUPD	%ZMM12,0x80(%RDI,%RAX,1)    [4]

0x43346e VMOVUPD	0xc0(%R11,%RAX,1),%ZMM13    [3]

0x433476 VMOVUPD	0xc0(%R9,%RAX,1),%ZMM15    [1]

0x43347e VADDPD	0xc0(%R10,%RAX,1),%ZMM13,%ZMM14    [2]

0x433486 VADDPD	0xc0(%R8,%RAX,1),%ZMM15,%ZMM0    [5]

0x43348e VADDPD	%ZMM0,%ZMM14,%ZMM5

0x433494 VMULPD	%ZMM1,%ZMM5,%ZMM2

0x43349a VMOVUPD	%ZMM2,0xc0(%RDI,%RAX,1)    [4]

0x4334a2 ADD	$0x100,%RAX

0x4334a8 CMP	%RAX,%R15

0x4334ab JNE	4333d7

/home/eoseret/qaas_runs_GNR/173-814-2343/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_mom.cpp: 88 - 88

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

88:           node_flux(i, j) = 0.25 * (mass_flux_x(i + 0, j - 1) + mass_flux_x(i, j) + mass_flux_x(i + 1, j - 1) + mass_flux_x(i + 1, j + 0));

Coverage (%)	Name	Source Location	Module
○97.64	gomp_thread_start	team.c:130	libgomp.so.1.0.0
○2.36	GOMP_parallel	libgomp.h:982	libgomp.so.1.0.0

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	1.26
Bottlenecks	P0, P1, P5,
Function	advec_mom_kernel(int, int, int, int, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer1D&, clover::Buffer1D&, int, int, int) [clone ._omp_fn.4]
Source	advec_mom.cpp:88-88
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	8.00
CQA cycles if no scalar integer	8.00
CQA cycles if FP arith vectorized	8.00
CQA cycles if fully vectorized	8.00
Front-end cycles	6.33
P0 cycles	8.00
P1 cycles	8.00
P2 cycles	5.33
P3 cycles	5.33
P4 cycles	2.00
P5 cycles	8.00
P6 cycles	1.00
P7 cycles	2.00
P8 cycles	2.00
P9 cycles	2.00
P10 cycles	0.40
P11 cycles	5.33
DIV/SQRT cycles	0.00
Inter-iter dependencies cycles	1
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	31.00
Nb uops	30.00
Nb loads	16.00
Nb stores	4.00
Nb stack references	0.00
FLOP/cycle	16.00
Nb FLOP add-sub	96.00
Nb FLOP mul	32.00
Nb FLOP fma	0.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	160.00
Bytes prefetched	0.00
Bytes loaded	1024.00
Bytes stored	256.00
Stride 0	0.00
Stride 1	5.00
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	100.00
Vectorization ratio mul	100.00
Vectorization ratio add_sub	100.00
Vectorization ratio fma	NA
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	100.00
Vector-efficiency ratio mul	100.00
Vector-efficiency ratio add_sub	100.00
Vector-efficiency ratio fma	NA
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	1.26
Bottlenecks	P0, P1, P5,
Function	advec_mom_kernel(int, int, int, int, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer1D&, clover::Buffer1D&, int, int, int) [clone ._omp_fn.4]
Source	advec_mom.cpp:88-88
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	8.00
CQA cycles if no scalar integer	8.00
CQA cycles if FP arith vectorized	8.00
CQA cycles if fully vectorized	8.00
Front-end cycles	6.33
P0 cycles	8.00
P1 cycles	8.00
P2 cycles	5.33
P3 cycles	5.33
P4 cycles	2.00
P5 cycles	8.00
P6 cycles	1.00
P7 cycles	2.00
P8 cycles	2.00
P9 cycles	2.00
P10 cycles	0.40
P11 cycles	5.33
DIV/SQRT cycles	0.00
Inter-iter dependencies cycles	1
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	31.00
Nb uops	30.00
Nb loads	16.00
Nb stores	4.00
Nb stack references	0.00
FLOP/cycle	16.00
Nb FLOP add-sub	96.00
Nb FLOP mul	32.00
Nb FLOP fma	0.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	160.00
Bytes prefetched	0.00
Bytes loaded	1024.00
Bytes stored	256.00
Stride 0	0.00
Stride 1	5.00
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	100.00
Vectorization ratio mul	100.00
Vectorization ratio add_sub	100.00
Vectorization ratio fma	NA
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	100.00
Vector-efficiency ratio mul	100.00
Vector-efficiency ratio add_sub	100.00
Vector-efficiency ratio fma	NA
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	advec_mom_kernel(int, int, int, int, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer1D&, clover::Buffer1D&, int, int, int) [clone ._omp_fn.4]
Source file and lines	advec_mom.cpp:88-88
Module	exec

The loop is defined in /home/eoseret/qaas_runs_GNR/173-814-2343/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_mom.cpp:88.

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

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

Performance is limited by execution of FP multiply or FMA (fused multiply-add) operations (the FP multiply/FMA unit is a bottleneck). By removing all these bottlenecks, you can lower the cost of an iteration from 8.00 to 6.33 cycles (1.26x speedup).

Workaround

Reduce the number of FP multiply/FMA instructions

FMA

Presence of both ADD/SUB and MUL operations.

Workaround

Recompile with march=sapphirerapids. CQA target is Granite_Rapids_ad (Intel(R) Xeon(R) Granite Rapids) but specialization flags are -march=graniterapids
Try to change order in which elements are evaluated (using parentheses) in arithmetic expressions containing both ADD/SUB and MUL operations to enable your compiler to generate FMA instructions wherever possible. For instance a + b*c is a valid FMA (MUL then ADD). However (a+b)* c cannot be translated into an FMA (ADD then MUL).

Type of elements and instruction set

16 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 128 FP arithmetical operations:

96: addition or subtraction
32: multiply

The binary loop is loading 1024 bytes (128 double precision FP elements). The binary loop is storing 256 bytes (32 double precision FP elements).

Arithmetic intensity

Arithmetic intensity is 0.10 FP operations per loaded or stored byte.

Unroll opportunity

Loop is potentially data access bound.

Workaround

Unroll your loop if trip count is significantly higher than target unroll factor and if some data references are common to consecutive iterations. This can be done manually. Or by recompiling with -funroll-loops and/or -floop-unroll-and-jam. Or with the unroll (resp. unroll_and_jam) directive on top of the inner (resp. surrounding) loop. You can enforce an unroll factor: #pragma GCC unroll N

General properties

nb instructions	31
nb uops	30
loop length	218
used x86 registers	7
used mmx registers	0
used xmm registers	0
used ymm registers	0
used zmm registers	14
nb stack references	0
ADD-SUB / MUL ratio	3.00

Front-end

ASSUMED MACRO FUSION FIT IN UOP CACHE

micro-operation queue	6.33 cycles
front end	6.33 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11
uops	8.00	0.60	5.33	5.33	2.00	8.00	1.00	2.00	2.00	2.00	0.40	5.33
cycles	8.00	8.00	5.33	5.33	2.00	8.00	1.00	2.00	2.00	2.00	0.40	5.33

Execution ports to units layout:

P0 (256 bits): VPU, BRU, ALU, DIV/SQRT
P1 (256 bits): ALU, VPU
P2 (512 bits): load
P3 (512 bits): load
P4 (512 bits): store data
P5 (512 bits): ALU, VPU
P6: ALU, BRU
P7: store address
P8: store address
P9 (512 bits): store data
P10: ALU
P11 (512 bits): load

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

Cycles summary

Front-end	6.33
Dispatch	8.00
Data deps.	1.00
Overall L1	8.00

Vectorization ratios

all	100%
load	100%
store	100%
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)

Vector efficiency ratios

all	100%
load	100%
store	100%
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)

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 8.00 cycles. At this rate:

66% of peak load performance is reached (128.00 out of 192.00 bytes loaded per cycle (GB/s @ 1GHz))
25% of peak store performance is reached (32.00 out of 128.00 bytes stored per cycle (GB/s @ 1GHz))

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the loop is: 4333d7

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	Latency	Recip. throughput	Vectorization
VMOVUPD (%R11,%RAX,1),%ZMM8	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VMOVUPD (%R9,%RAX,1),%ZMM10	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VADDPD (%R10,%RAX,1),%ZMM8,%ZMM9	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD (%R8,%RAX,1),%ZMM10,%ZMM11	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD %ZMM11,%ZMM9,%ZMM12	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	3	0.50	vect (100.0%)
VMULPD %ZMM1,%ZMM12,%ZMM13	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	4	0.50	vect (100.0%)
VMOVUPD %ZMM13,(%RDI,%RAX,1)	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0-1	1	vect (100.0%)
VMOVUPD 0x40(%R11,%RAX,1),%ZMM14	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VMOVUPD 0x40(%R9,%RAX,1),%ZMM0	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VADDPD 0x40(%R10,%RAX,1),%ZMM14,%ZMM15	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD 0x40(%R8,%RAX,1),%ZMM0,%ZMM5	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD %ZMM5,%ZMM15,%ZMM2	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	3	0.50	vect (100.0%)
VMULPD %ZMM1,%ZMM2,%ZMM6	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	4	0.50	vect (100.0%)
VMOVUPD %ZMM6,0x40(%RDI,%RAX,1)	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0-1	1	vect (100.0%)
VMOVUPD 0x80(%R11,%RAX,1),%ZMM7	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VMOVUPD 0x80(%R9,%RAX,1),%ZMM9	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VADDPD 0x80(%R10,%RAX,1),%ZMM7,%ZMM8	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD 0x80(%R8,%RAX,1),%ZMM9,%ZMM10	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD %ZMM10,%ZMM8,%ZMM11	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	3	0.50	vect (100.0%)
VMULPD %ZMM1,%ZMM11,%ZMM12	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	4	0.50	vect (100.0%)
VMOVUPD %ZMM12,0x80(%RDI,%RAX,1)	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0-1	1	vect (100.0%)
VMOVUPD 0xc0(%R11,%RAX,1),%ZMM13	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VMOVUPD 0xc0(%R9,%RAX,1),%ZMM15	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VADDPD 0xc0(%R10,%RAX,1),%ZMM13,%ZMM14	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD 0xc0(%R8,%RAX,1),%ZMM15,%ZMM0	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD %ZMM0,%ZMM14,%ZMM5	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	3	0.50	vect (100.0%)
VMULPD %ZMM1,%ZMM5,%ZMM2	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	4	0.50	vect (100.0%)
VMOVUPD %ZMM2,0xc0(%RDI,%RAX,1)	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0-1	1	vect (100.0%)
ADD $0x100,%RAX	1	0.20	0.20	0	0	0	0.20	0.20	0	0	0	0.20	0	1	0.20	N/A
CMP %RAX,%R15	1	0.20	0.20	0	0	0	0.20	0.20	0	0	0	0.20	0	1	0.20	N/A
JNE 4333d7 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.4+0x237>	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0	0.50	N/A

Function	advec_mom_kernel(int, int, int, int, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer1D&, clover::Buffer1D&, int, int, int) [clone ._omp_fn.4]
Source file and lines	advec_mom.cpp:88-88
Module	exec

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

Workaround

Reduce the number of FP multiply/FMA instructions

FMA

Presence of both ADD/SUB and MUL operations.

Workaround

Recompile with march=sapphirerapids. CQA target is Granite_Rapids_ad (Intel(R) Xeon(R) Granite Rapids) but specialization flags are -march=graniterapids
Try to change order in which elements are evaluated (using parentheses) in arithmetic expressions containing both ADD/SUB and MUL operations to enable your compiler to generate FMA instructions wherever possible. For instance a + b*c is a valid FMA (MUL then ADD). However (a+b)* c cannot be translated into an FMA (ADD then MUL).

Type of elements and instruction set

16 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 128 FP arithmetical operations:

96: addition or subtraction
32: multiply

The binary loop is loading 1024 bytes (128 double precision FP elements). The binary loop is storing 256 bytes (32 double precision FP elements).

Arithmetic intensity

Arithmetic intensity is 0.10 FP operations per loaded or stored byte.

Unroll opportunity

Loop is potentially data access bound.

Workaround

General properties

nb instructions	31
nb uops	30
loop length	218
used x86 registers	7
used mmx registers	0
used xmm registers	0
used ymm registers	0
used zmm registers	14
nb stack references	0
ADD-SUB / MUL ratio	3.00

Front-end

ASSUMED MACRO FUSION FIT IN UOP CACHE

micro-operation queue	6.33 cycles
front end	6.33 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11
uops	8.00	0.60	5.33	5.33	2.00	8.00	1.00	2.00	2.00	2.00	0.40	5.33
cycles	8.00	8.00	5.33	5.33	2.00	8.00	1.00	2.00	2.00	2.00	0.40	5.33

Execution ports to units layout:

P0 (256 bits): VPU, BRU, ALU, DIV/SQRT
P1 (256 bits): ALU, VPU
P2 (512 bits): load
P3 (512 bits): load
P4 (512 bits): store data
P5 (512 bits): ALU, VPU
P6: ALU, BRU
P7: store address
P8: store address
P9 (512 bits): store data
P10: ALU
P11 (512 bits): load

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

Cycles summary

Front-end	6.33
Dispatch	8.00
Data deps.	1.00
Overall L1	8.00

Vectorization ratios

all	100%
load	100%
store	100%
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)

Vector efficiency ratios

all	100%
load	100%
store	100%
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)

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 8.00 cycles. At this rate:

66% of peak load performance is reached (128.00 out of 192.00 bytes loaded per cycle (GB/s @ 1GHz))
25% of peak store performance is reached (32.00 out of 128.00 bytes stored per cycle (GB/s @ 1GHz))

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the loop is: 4333d7

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	Latency	Recip. throughput	Vectorization
VMOVUPD (%R11,%RAX,1),%ZMM8	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VMOVUPD (%R9,%RAX,1),%ZMM10	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VADDPD (%R10,%RAX,1),%ZMM8,%ZMM9	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD (%R8,%RAX,1),%ZMM10,%ZMM11	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD %ZMM11,%ZMM9,%ZMM12	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	3	0.50	vect (100.0%)
VMULPD %ZMM1,%ZMM12,%ZMM13	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	4	0.50	vect (100.0%)
VMOVUPD %ZMM13,(%RDI,%RAX,1)	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0-1	1	vect (100.0%)
VMOVUPD 0x40(%R11,%RAX,1),%ZMM14	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VMOVUPD 0x40(%R9,%RAX,1),%ZMM0	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VADDPD 0x40(%R10,%RAX,1),%ZMM14,%ZMM15	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD 0x40(%R8,%RAX,1),%ZMM0,%ZMM5	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD %ZMM5,%ZMM15,%ZMM2	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	3	0.50	vect (100.0%)
VMULPD %ZMM1,%ZMM2,%ZMM6	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	4	0.50	vect (100.0%)
VMOVUPD %ZMM6,0x40(%RDI,%RAX,1)	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0-1	1	vect (100.0%)
VMOVUPD 0x80(%R11,%RAX,1),%ZMM7	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VMOVUPD 0x80(%R9,%RAX,1),%ZMM9	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VADDPD 0x80(%R10,%RAX,1),%ZMM7,%ZMM8	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD 0x80(%R8,%RAX,1),%ZMM9,%ZMM10	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD %ZMM10,%ZMM8,%ZMM11	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	3	0.50	vect (100.0%)
VMULPD %ZMM1,%ZMM11,%ZMM12	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	4	0.50	vect (100.0%)
VMOVUPD %ZMM12,0x80(%RDI,%RAX,1)	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0-1	1	vect (100.0%)
VMOVUPD 0xc0(%R11,%RAX,1),%ZMM13	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VMOVUPD 0xc0(%R9,%RAX,1),%ZMM15	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	0-1	0.50	vect (100.0%)
VADDPD 0xc0(%R10,%RAX,1),%ZMM13,%ZMM14	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD 0xc0(%R8,%RAX,1),%ZMM15,%ZMM0	1	0.50	0	0.33	0.33	0	0.50	0	0	0	0	0	0.33	3	0.65	vect (100.0%)
VADDPD %ZMM0,%ZMM14,%ZMM5	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	3	0.50	vect (100.0%)
VMULPD %ZMM1,%ZMM5,%ZMM2	1	0.50	0	0	0	0	0.50	0	0	0	0	0	0	4	0.50	vect (100.0%)
VMOVUPD %ZMM2,0xc0(%RDI,%RAX,1)	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0-1	1	vect (100.0%)
ADD $0x100,%RAX	1	0.20	0.20	0	0	0	0.20	0.20	0	0	0	0.20	0	1	0.20	N/A
CMP %RAX,%R15	1	0.20	0.20	0	0	0	0.20	0.20	0	0	0	0.20	0	1	0.20	N/A
JNE 4333d7 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.4+0x237>	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0	0.50	N/A

Report Configuration

Vectorization

Details

Execution units bottlenecks

Workaround

FMA

Workaround

Type of elements and instruction set

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

Arithmetic intensity

Unroll opportunity

Workaround

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

Workaround

FMA

Workaround

Type of elements and instruction set

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

Arithmetic intensity

Unroll opportunity

Workaround

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code