OV - exec - Function main._omp

	Function: main._omp_fn.3	Module: exec	Source: stream.c:290-292	Coverage (incl. loops): 0.67% \| (excl. loops): 0.00%

/beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c: 290 - 292

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

290: #pragma omp parallel for

291:     for (j = 0; j < STREAM_ARRAY_SIZE; j++)

292: 		a[j] = 2.0E0 * a[j];

0x4023c0 PUSH	%RBP

0x4023c1 MOV	$0x401060,%RAX

0x4023cb MOV	%RSP,%RBP

0x4023ce PUSH	%RBX

0x4023cf AND	$-0x40,%RSP

0x4023d3 CALL	%RAX

0x4023d5 MOVSXD	%EAX,%RBX

0x4023d8 MOV	$0x401030,%RAX

0x4023e2 CALL	%RAX

0x4023e4 MOVSXD	%EAX,%RCX

0x4023e7 MOV	$0x33450000,%EAX

0x4023ec CQTO

0x4023ee IDIV	%RBX

0x4023f1 CMP	%RDX,%RCX

0x4023f4 JL	402670

0x4023fa IMUL	%RAX,%RCX

0x4023fe ADD	%RCX,%RDX

0x402401 LEA	(%RAX,%RDX,1),%R8

0x402405 CMP	%R8,%RDX

0x402408 JGE	402654

0x40240e LEA	-0x1(%RAX),%RSI

0x402412 MOV	%RDX,%R9

0x402415 CMP	$0x6,%RSI

0x402419 JBE	402680

0x40241f MOV	%RAX,%RBX

0x402422 MOV	$0x334906100,%RDI

0x40242c SHR	$0x3,%RBX

0x402430 SAL	$0x6,%RBX

0x402434 LEA	(%RDI,%RDX,8),%R11

0x402438 LEA	(%RBX,%R11,1),%RCX

0x40243c SUB	$0x40,%RBX

0x402440 SHR	$0x6,%RBX

0x402444 INC	%RBX

0x402447 AND	$0x7,%EBX

0x40244a JE	402526

0x402450 CMP	$0x1,%RBX

0x402454 JE	402506

0x40245a CMP	$0x2,%RBX

0x40245e JE	4024ef

0x402464 CMP	$0x3,%RBX

0x402468 JE	4024d8

0x40246a CMP	$0x4,%RBX

0x40246e JE	4024c1

0x402470 CMP	$0x5,%RBX

0x402474 JE	4024aa

0x402476 CMP	$0x6,%RBX

0x40247a JE	402493

0x40247c VMOVUPD	(%R11),%ZMM0

0x402482 ADD	$0x40,%R11

0x402486 VADDPD	%ZMM0,%ZMM0,%ZMM1

0x40248c VMOVUPD	%ZMM1,-0x40(%R11)

0x402493 VMOVUPD	(%R11),%ZMM2

0x402499 ADD	$0x40,%R11

0x40249d VADDPD	%ZMM2,%ZMM2,%ZMM3

0x4024a3 VMOVUPD	%ZMM3,-0x40(%R11)

0x4024aa VMOVUPD	(%R11),%ZMM4

0x4024b0 ADD	$0x40,%R11

0x4024b4 VADDPD	%ZMM4,%ZMM4,%ZMM5

0x4024ba VMOVUPD	%ZMM5,-0x40(%R11)

0x4024c1 VMOVUPD	(%R11),%ZMM6

0x4024c7 ADD	$0x40,%R11

0x4024cb VADDPD	%ZMM6,%ZMM6,%ZMM7

0x4024d1 VMOVUPD	%ZMM7,-0x40(%R11)

0x4024d8 VMOVUPD	(%R11),%ZMM8

0x4024de ADD	$0x40,%R11

0x4024e2 VADDPD	%ZMM8,%ZMM8,%ZMM9

0x4024e8 VMOVUPD	%ZMM9,-0x40(%R11)

0x4024ef VMOVUPD	(%R11),%ZMM10

0x4024f5 ADD	$0x40,%R11

0x4024f9 VADDPD	%ZMM10,%ZMM10,%ZMM11

0x4024ff VMOVUPD	%ZMM11,-0x40(%R11)

0x402506 VMOVUPD	(%R11),%ZMM12

0x40250c ADD	$0x40,%R11

0x402510 VADDPD	%ZMM12,%ZMM12,%ZMM13

0x402516 VMOVUPD	%ZMM13,-0x40(%R11)

0x40251d CMP	%RCX,%R11

0x402520 JE	4025d5

(3) 0x402526 VMOVUPD	(%R11),%ZMM14

(3) 0x40252c VMOVUPD	0x40(%R11),%ZMM0

(3) 0x402533 VMOVUPD	0x80(%R11),%ZMM2

(3) 0x40253a ADD	$0x200,%R11

(3) 0x402541 VMOVUPD	-0x140(%R11),%ZMM4

(3) 0x402548 VMOVUPD	-0x100(%R11),%ZMM6

(3) 0x40254f VMOVUPD	-0xc0(%R11),%ZMM8

(3) 0x402556 VMOVUPD	-0x80(%R11),%ZMM10

(3) 0x40255d VMOVUPD	-0x40(%R11),%ZMM12

(3) 0x402564 VADDPD	%ZMM14,%ZMM14,%ZMM15

(3) 0x40256a VADDPD	%ZMM0,%ZMM0,%ZMM1

(3) 0x402570 VADDPD	%ZMM2,%ZMM2,%ZMM3

(3) 0x402576 VADDPD	%ZMM4,%ZMM4,%ZMM5

(3) 0x40257c VMOVUPD	%ZMM15,-0x200(%R11)

(3) 0x402583 VADDPD	%ZMM6,%ZMM6,%ZMM7

(3) 0x402589 VADDPD	%ZMM8,%ZMM8,%ZMM9

(3) 0x40258f VMOVUPD	%ZMM1,-0x1c0(%R11)

(3) 0x402596 VMOVUPD	%ZMM3,-0x180(%R11)

(3) 0x40259d VADDPD	%ZMM10,%ZMM10,%ZMM11

(3) 0x4025a3 VADDPD	%ZMM12,%ZMM12,%ZMM13

(3) 0x4025a9 VMOVUPD	%ZMM5,-0x140(%R11)

(3) 0x4025b0 VMOVUPD	%ZMM7,-0x100(%R11)

(3) 0x4025b7 VMOVUPD	%ZMM9,-0xc0(%R11)

(3) 0x4025be VMOVUPD	%ZMM11,-0x80(%R11)

(3) 0x4025c5 VMOVUPD	%ZMM13,-0x40(%R11)

(3) 0x4025cc CMP	%RCX,%R11

(3) 0x4025cf JNE	402526

0x4025d5 MOV	%RAX,%R10

0x4025d8 AND	$-0x8,%R10

0x4025dc ADD	%R10,%RDX

0x4025df CMP	%R10,%RAX

0x4025e2 JE	402660

0x4025e4 SUB	%R10,%RAX

0x4025e7 LEA	-0x1(%RAX),%RSI

0x4025eb CMP	$0x2,%RSI

0x4025ef JBE	402615

0x4025f1 ADD	%R9,%R10

0x4025f4 LEA	(%RDI,%R10,8),%R9

0x4025f8 MOV	%RAX,%R10

0x4025fb VMOVUPD	(%R9),%YMM14

0x402600 AND	$-0x4,%R10

0x402604 ADD	%R10,%RDX

0x402607 VADDPD	%YMM14,%YMM14,%YMM15

0x40260c VMOVUPD	%YMM15,(%R9)

0x402611 TEST	$0x3,%AL

0x402613 JE	402660

0x402615 LEA	0x1(%RDX),%RAX

0x402619 VMOVSD	(%RDI,%RDX,8),%XMM0

0x40261e VADDSD	%XMM0,%XMM0,%XMM1

0x402622 VMOVSD	%XMM1,(%RDI,%RDX,8)

0x402627 CMP	%RAX,%R8

0x40262a JLE	402660

0x40262c ADD	$0x2,%RDX

0x402630 VMOVSD	(%RDI,%RAX,8),%XMM2

0x402635 VADDSD	%XMM2,%XMM2,%XMM3

0x402639 VMOVSD	%XMM3,(%RDI,%RAX,8)

0x40263e CMP	%RDX,%R8

0x402641 JLE	402660

0x402643 VMOVSD	(%RDI,%RDX,8),%XMM4

0x402648 VADDSD	%XMM4,%XMM4,%XMM5

0x40264c VMOVSD	%XMM5,(%RDI,%RDX,8)

0x402651 VZEROUPPER

0x402654 MOV	-0x8(%RBP),%RBX

0x402658 LEAVE

0x402659 RET

0x40265a NOPW	(%RAX,%RAX,1)

0x402660 VZEROUPPER

0x402663 MOV	-0x8(%RBP),%RBX

0x402667 LEAVE

0x402668 RET

0x402669 NOPL	(%RAX)

0x402670 INC	%RAX

0x402673 XOR	%EDX,%EDX

0x402675 JMP	4023fa

0x40267a NOPW	(%RAX,%RAX,1)

0x402680 XOR	%R10D,%R10D

0x402683 MOV	$0x334906100,%RDI

0x40268d JMP	4025e4

0x402692 NOPW	%CS:(%RAX,%RAX,1)

0x40269d NOPL	(%RAX)

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

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 /

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

The code analyzed by CQA in that panel excludes loops and represents 0.00% of application time for run gcc_0

Source file and lines	stream.c:290-292
Module	exec

The function is defined in /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:290-292.

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 22.67 to 10.83 cycles (2.09x 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 48% of vector register length is used (average across all SSE/AVX instructions). By fully vectorizing your function, you can lower the cost of an iteration from 22.67 to 18.02 cycles (1.26x speedup).

Details

48% of SSE/AVX instructions are used in vector version (process two or more data elements in vector registers):

72% of SSE/AVX loads are used in vector version.
72% of SSE/AVX stores are used in vector version.
72% of SSE/AVX addition or subtraction instructions are used in vector version.
0% of SSE/AVX multiply instructions are used in vector version.
0% of SSE/AVX divide and square root instructions are used in vector version.
10% of SSE/AVX 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)

Execution units bottlenecks

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

No data for this section

CALL instructions

Detected function call instructions.

Details

Calling (and then returning from) a function prevents many compiler optimizations (like vectorization), breaks control flow (which reduces pipeline performance) and executes extra instructions to save/restore the registers used inside it, which is very expensive (dozens of cycles). Consider to inline small functions.

unknown: 2 occurrences►
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:290
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:290

Complex instructions

Detected COMPLEX INSTRUCTIONS.

Details

These instructions generate more than one micro-operation and only one of them can be decoded during a cycle and the extra micro-operations increase pressure on execution units.

IDIV: 1 occurrences►
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:290
LEAVE: 2 occurrences►
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:290
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:290
VMOVUPD: 7 occurrences►
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292

Vector unaligned load/store instructions

Detected 2 optimal vector unaligned load/store instructions.

Details

VMOVUPD: 2 occurrences►
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292

Workaround

Use vector aligned instructions:

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

Conversion instructions

Detected expensive conversion instructions.

Details

CQTO: 1 occurrences►
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:290

Workaround

Avoid mixing data with different types. In particular, check if the type of constants is the same as array elements.

Type of elements and instruction set

3 SSE or AVX instructions are processing arithmetic or math operations on double precision FP elements in scalar mode (one at a time). 1 AVX instructions are processing arithmetic or math operations on double precision FP elements in vector mode (four at a time). 7 AVX-512 instructions are processing arithmetic or math operations on double precision FP elements in vector mode (eight at a time).

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

The binary function is composed of 63 FP arithmetical operations:

63: addition or subtraction

The binary function is loading 520 bytes (65 double precision FP elements). The binary function is storing 504 bytes (63 double precision FP elements).

Arithmetic intensity

Arithmetic intensity is 0.06 FP operations per loaded or stored byte.

General properties

nb instructions	130
nb uops	136
loop length	561
used x86 registers	12
used mmx registers	0
used xmm registers	6
used ymm registers	2
used zmm registers	14
nb stack references	1

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	22.67 cycles
front end	22.67 cycles

Back-end

	ALU0/BRU0	ALU1	ALU2	ALU3	BRU1	AGU0	AGU1	AGU2	FP0	FP1	FP2	FP3	FP4	FP5
uops	16.75	16.75	16.75	16.75	12.00	9.33	9.33	9.33	0.00	0.00	5.50	5.50	9.00	9.00
cycles	16.75	16.75	16.75	16.75	12.00	11.67	11.67	11.67	0.00	0.00	9.00	9.00	9.00	9.00

Execution ports to units layout:

ALU0/BRU0: ALU, BRU
ALU1: ALU
ALU2: ALU
ALU3: ALU
BRU1: BRU
AGU0 (256 bits): store address, load
AGU1 (256 bits): store address, load
AGU2 (256 bits): store address, load
FP0 (256 bits): VPU, DIV/SQRT
FP1 (256 bits): VPU, DIV/SQRT
FP2 (256 bits): VPU
FP3 (256 bits): VPU
FP4 (256 bits): FP store data
FP5 (256 bits): FP store data

Cycles executing div or sqrt instructions

7.00-12.00

Cycles summary

Front-end	22.67
Dispatch	16.75
DIV/SQRT	7.00-12.00
Overall L1	22.67

Vectorization ratios

INT

all	9%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	0%
add-sub	NA (no add-sub vectorizable/vectorized instructions)
fma	NA (no fma vectorizable/vectorized instructions)
other	10%

all	72%
load	72%
store	72%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	72%
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	48%
load	72%
store	72%
mul	0%
add-sub	72%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	0%
other	10%

Vector efficiency ratios

INT

all	13%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	12%
add-sub	NA (no add-sub vectorizable/vectorized instructions)
fma	NA (no fma vectorizable/vectorized instructions)
other	13%

all	71%
load	71%
store	71%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	71%
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	48%
load	71%
store	71%
mul	12%
add-sub	71%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	12%
other	13%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 22.67 cycles. At this rate:

23% of peak load performance is reached (22.94 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
34% of peak store performance is reached (22.24 out of 64.00 bytes stored per cycle (GB/s @ 1GHz))

Front-end bottlenecks

Performance is limited by instruction throughput (loading/decoding program instructions to execution core) (front-end is a bottleneck). By removing all these bottlenecks, you can lower the cost of an iteration from 22.67 to 16.75 cycles (1.35x speedup).

ASM code

In the binary file, the address of the function is: 4023c0

Instruction	Nb FU	ALU0/BRU0	ALU1	ALU2	ALU3	BRU1	AGU0	AGU1	AGU2	FP0	FP1	FP2	FP3	FP4	FP5	Latency	Recip. throughput	Vectorization
PUSH %RBP	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
MOV $0x401060,%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
MOV %RSP,%RBP	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
PUSH %RBX	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
AND $-0x40,%RSP	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
CALL %RAX	2	0.50	0	0	0	0.50	0.33	0.33	0.33	0	0	0	0	0	0	0	2	N/A
MOVSXD %EAX,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
MOV $0x401030,%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
CALL %RAX	2	0.50	0	0	0	0.50	0.33	0.33	0.33	0	0	0	0	0	0	0	2	N/A
MOVSXD %EAX,%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (6.3%)
MOV $0x33450000,%EAX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
CQTO																		scal (12.5%)
IDIV %RBX	2	2	0	0	0	0	0	0	0	0	0	0	0	0	0	9-19	7-12	scal (12.5%)
CMP %RDX,%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JL 402670 <main._omp_fn.3+0x2b0>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
IMUL %RAX,%RCX	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	3	1	scal (12.5%)
ADD %RCX,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
LEA (%RAX,%RDX,1),%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
CMP %R8,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JGE 402654 <main._omp_fn.3+0x294>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
LEA -0x1(%RAX),%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
MOV %RDX,%R9	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
CMP $0x6,%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JBE 402680 <main._omp_fn.3+0x2c0>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
MOV %RAX,%RBX	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV $0x334906100,%RDI	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
SHR $0x3,%RBX	1	0	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SAL $0x6,%RBX	1	0	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LEA (%RDI,%RDX,8),%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
LEA (%RBX,%R11,1),%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
SUB $0x40,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
SHR $0x6,%RBX	1	0	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
INC %RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
AND $0x7,%EBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
JE 402526 <main._omp_fn.3+0x166>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
CMP $0x1,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 402506 <main._omp_fn.3+0x146>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
CMP $0x2,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 4024ef <main._omp_fn.3+0x12f>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
CMP $0x3,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 4024d8 <main._omp_fn.3+0x118>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
CMP $0x4,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 4024c1 <main._omp_fn.3+0x101>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
CMP $0x5,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 4024aa <main._omp_fn.3+0xea>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
CMP $0x6,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 402493 <main._omp_fn.3+0xd3>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
VMOVUPD (%R11),%ZMM0	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	1	vect (100.0%)
ADD $0x40,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %ZMM0,%ZMM0,%ZMM1	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	1	vect (100.0%)
VMOVUPD %ZMM1,-0x40(%R11)	2	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	1	1	5	2	vect (100.0%)
VMOVUPD (%R11),%ZMM2	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	1	vect (100.0%)
ADD $0x40,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %ZMM2,%ZMM2,%ZMM3	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	1	vect (100.0%)
VMOVUPD %ZMM3,-0x40(%R11)	2	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	1	1	5	2	vect (100.0%)
VMOVUPD (%R11),%ZMM4	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	1	vect (100.0%)
ADD $0x40,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %ZMM4,%ZMM4,%ZMM5	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	1	vect (100.0%)
VMOVUPD %ZMM5,-0x40(%R11)	2	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	1	1	5	2	vect (100.0%)
VMOVUPD (%R11),%ZMM6	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	1	vect (100.0%)
ADD $0x40,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %ZMM6,%ZMM6,%ZMM7	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	1	vect (100.0%)
VMOVUPD %ZMM7,-0x40(%R11)	2	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	1	1	5	2	vect (100.0%)
VMOVUPD (%R11),%ZMM8	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	1	vect (100.0%)
ADD $0x40,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %ZMM8,%ZMM8,%ZMM9	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	1	vect (100.0%)
VMOVUPD %ZMM9,-0x40(%R11)	2	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	1	1	5	2	vect (100.0%)
VMOVUPD (%R11),%ZMM10	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	1	vect (100.0%)
ADD $0x40,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %ZMM10,%ZMM10,%ZMM11	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	1	vect (100.0%)
VMOVUPD %ZMM11,-0x40(%R11)	2	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	1	1	5	2	vect (100.0%)
VMOVUPD (%R11),%ZMM12	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	1	vect (100.0%)
ADD $0x40,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %ZMM12,%ZMM12,%ZMM13	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	1	vect (100.0%)
VMOVUPD %ZMM13,-0x40(%R11)	2	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	1	1	5	2	vect (100.0%)
CMP %RCX,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 4025d5 <main._omp_fn.3+0x215>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
MOV %RAX,%R10	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
AND $-0x8,%R10	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
ADD %R10,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
CMP %R10,%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 402660 <main._omp_fn.3+0x2a0>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
SUB %R10,%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
LEA -0x1(%RAX),%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
CMP $0x2,%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JBE 402615 <main._omp_fn.3+0x255>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
ADD %R9,%R10	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
LEA (%RDI,%R10,8),%R9	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
MOV %RAX,%R10	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
VMOVUPD (%R9),%YMM14	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	0.50	vect (50.0%)
AND $-0x4,%R10	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
ADD %R10,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %YMM14,%YMM14,%YMM15	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	vect (50.0%)
VMOVUPD %YMM15,(%R9)	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0.50	0.50	4	1	vect (50.0%)
TEST $0x3,%AL	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
JE 402660 <main._omp_fn.3+0x2a0>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
LEA 0x1(%RDX),%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VMOVSD (%RDI,%RDX,8),%XMM0	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50	scal (12.5%)
VADDSD %XMM0,%XMM0,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	scal (12.5%)
VMOVSD %XMM1,(%RDI,%RDX,8)	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0.50	0.50	1	1	scal (12.5%)
CMP %RAX,%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JLE 402660 <main._omp_fn.3+0x2a0>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
ADD $0x2,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VMOVSD (%RDI,%RAX,8),%XMM2	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50	scal (12.5%)
VADDSD %XMM2,%XMM2,%XMM3	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	scal (12.5%)
VMOVSD %XMM3,(%RDI,%RAX,8)	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0.50	0.50	1	1	scal (12.5%)
CMP %RDX,%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JLE 402660 <main._omp_fn.3+0x2a0>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
VMOVSD (%RDI,%RDX,8),%XMM4	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50	scal (12.5%)
VADDSD %XMM4,%XMM4,%XMM5	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	scal (12.5%)
VMOVSD %XMM5,(%RDI,%RDX,8)	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0.50	0.50	1	1	scal (12.5%)
VZEROUPPER	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	vect (25.0%)
MOV -0x8(%RBP),%RBX	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	0.33	N/A
LEAVE	2	0.25	0.25	0.25	0.25	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50	N/A
RET	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50	N/A
NOPW (%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
VZEROUPPER	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	vect (25.0%)
MOV -0x8(%RBP),%RBX	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	0.33	N/A
LEAVE	2	0.25	0.25	0.25	0.25	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50	N/A
RET	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50	N/A
NOPL (%RAX)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
INC %RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
XOR %EDX,%EDX	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.25	N/A
JMP 4023fa <main._omp_fn.3+0x3a>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	0	1	N/A
NOPW (%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
XOR %R10D,%R10D	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.25	N/A
MOV $0x334906100,%RDI	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
JMP 4025e4 <main._omp_fn.3+0x224>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	0	1	N/A
NOPW %CS:(%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
NOPL (%RAX)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A

The code analyzed by CQA in that panel excludes loops and represents 0.00% of application time for run gcc_0

Source file and lines	stream.c:290-292
Module	exec

gain
potential
hint
expert

Code clean check

Workaround

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

Vectorization

Details

48% of SSE/AVX instructions are used in vector version (process two or more data elements in vector registers):

72% of SSE/AVX loads are used in vector version.
72% of SSE/AVX stores are used in vector version.
72% of SSE/AVX addition or subtraction instructions are used in vector version.
0% of SSE/AVX multiply instructions are used in vector version.
0% of SSE/AVX divide and square root instructions are used in vector version.
10% of SSE/AVX 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)

Execution units bottlenecks

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

No data for this section

CALL instructions

Detected function call instructions.

Details

unknown: 2 occurrences►
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:290
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:290

Complex instructions

Detected COMPLEX INSTRUCTIONS.

Details

These instructions generate more than one micro-operation and only one of them can be decoded during a cycle and the extra micro-operations increase pressure on execution units.

IDIV: 1 occurrences►
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:290
LEAVE: 2 occurrences►
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:290
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:290
VMOVUPD: 7 occurrences►
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292

Vector unaligned load/store instructions

Detected 2 optimal vector unaligned load/store instructions.

Details

VMOVUPD: 2 occurrences►
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:292

Workaround

Use vector aligned instructions:

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

Conversion instructions

Detected expensive conversion instructions.

Details

CQTO: 1 occurrences►
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:290

Workaround

Avoid mixing data with different types. In particular, check if the type of constants is the same as array elements.

Type of elements and instruction set

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

The binary function is composed of 63 FP arithmetical operations:

63: addition or subtraction

The binary function is loading 520 bytes (65 double precision FP elements). The binary function is storing 504 bytes (63 double precision FP elements).

Arithmetic intensity

Arithmetic intensity is 0.06 FP operations per loaded or stored byte.

General properties

nb instructions	130
nb uops	136
loop length	561
used x86 registers	12
used mmx registers	0
used xmm registers	6
used ymm registers	2
used zmm registers	14
nb stack references	1

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	22.67 cycles
front end	22.67 cycles

Back-end

	ALU0/BRU0	ALU1	ALU2	ALU3	BRU1	AGU0	AGU1	AGU2	FP0	FP1	FP2	FP3	FP4	FP5
uops	16.75	16.75	16.75	16.75	12.00	9.33	9.33	9.33	0.00	0.00	5.50	5.50	9.00	9.00
cycles	16.75	16.75	16.75	16.75	12.00	11.67	11.67	11.67	0.00	0.00	9.00	9.00	9.00	9.00

Execution ports to units layout:

ALU0/BRU0: ALU, BRU
ALU1: ALU
ALU2: ALU
ALU3: ALU
BRU1: BRU
AGU0 (256 bits): store address, load
AGU1 (256 bits): store address, load
AGU2 (256 bits): store address, load
FP0 (256 bits): VPU, DIV/SQRT
FP1 (256 bits): VPU, DIV/SQRT
FP2 (256 bits): VPU
FP3 (256 bits): VPU
FP4 (256 bits): FP store data
FP5 (256 bits): FP store data

Cycles executing div or sqrt instructions

7.00-12.00

Cycles summary

Front-end	22.67
Dispatch	16.75
DIV/SQRT	7.00-12.00
Overall L1	22.67

Vectorization ratios

INT

all	9%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	0%
add-sub	NA (no add-sub vectorizable/vectorized instructions)
fma	NA (no fma vectorizable/vectorized instructions)
other	10%

all	72%
load	72%
store	72%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	72%
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	48%
load	72%
store	72%
mul	0%
add-sub	72%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	0%
other	10%

Vector efficiency ratios

INT

all	13%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	12%
add-sub	NA (no add-sub vectorizable/vectorized instructions)
fma	NA (no fma vectorizable/vectorized instructions)
other	13%

all	71%
load	71%
store	71%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	71%
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	48%
load	71%
store	71%
mul	12%
add-sub	71%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	12%
other	13%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 22.67 cycles. At this rate:

23% of peak load performance is reached (22.94 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
34% of peak store performance is reached (22.24 out of 64.00 bytes stored per cycle (GB/s @ 1GHz))

Front-end bottlenecks

ASM code

In the binary file, the address of the function is: 4023c0

Instruction	Nb FU	ALU0/BRU0	ALU1	ALU2	ALU3	BRU1	AGU0	AGU1	AGU2	FP0	FP1	FP2	FP3	FP4	FP5	Latency	Recip. throughput	Vectorization
PUSH %RBP	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
MOV $0x401060,%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
MOV %RSP,%RBP	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
PUSH %RBX	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
AND $-0x40,%RSP	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
CALL %RAX	2	0.50	0	0	0	0.50	0.33	0.33	0.33	0	0	0	0	0	0	0	2	N/A
MOVSXD %EAX,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
MOV $0x401030,%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
CALL %RAX	2	0.50	0	0	0	0.50	0.33	0.33	0.33	0	0	0	0	0	0	0	2	N/A
MOVSXD %EAX,%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (6.3%)
MOV $0x33450000,%EAX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
CQTO																		scal (12.5%)
IDIV %RBX	2	2	0	0	0	0	0	0	0	0	0	0	0	0	0	9-19	7-12	scal (12.5%)
CMP %RDX,%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JL 402670 <main._omp_fn.3+0x2b0>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
IMUL %RAX,%RCX	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	3	1	scal (12.5%)
ADD %RCX,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
LEA (%RAX,%RDX,1),%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
CMP %R8,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JGE 402654 <main._omp_fn.3+0x294>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
LEA -0x1(%RAX),%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
MOV %RDX,%R9	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
CMP $0x6,%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JBE 402680 <main._omp_fn.3+0x2c0>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
MOV %RAX,%RBX	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV $0x334906100,%RDI	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
SHR $0x3,%RBX	1	0	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SAL $0x6,%RBX	1	0	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LEA (%RDI,%RDX,8),%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
LEA (%RBX,%R11,1),%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
SUB $0x40,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
SHR $0x6,%RBX	1	0	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
INC %RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
AND $0x7,%EBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
JE 402526 <main._omp_fn.3+0x166>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
CMP $0x1,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 402506 <main._omp_fn.3+0x146>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
CMP $0x2,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 4024ef <main._omp_fn.3+0x12f>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
CMP $0x3,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 4024d8 <main._omp_fn.3+0x118>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
CMP $0x4,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 4024c1 <main._omp_fn.3+0x101>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
CMP $0x5,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 4024aa <main._omp_fn.3+0xea>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
CMP $0x6,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 402493 <main._omp_fn.3+0xd3>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
VMOVUPD (%R11),%ZMM0	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	1	vect (100.0%)
ADD $0x40,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %ZMM0,%ZMM0,%ZMM1	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	1	vect (100.0%)
VMOVUPD %ZMM1,-0x40(%R11)	2	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	1	1	5	2	vect (100.0%)
VMOVUPD (%R11),%ZMM2	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	1	vect (100.0%)
ADD $0x40,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %ZMM2,%ZMM2,%ZMM3	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	1	vect (100.0%)
VMOVUPD %ZMM3,-0x40(%R11)	2	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	1	1	5	2	vect (100.0%)
VMOVUPD (%R11),%ZMM4	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	1	vect (100.0%)
ADD $0x40,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %ZMM4,%ZMM4,%ZMM5	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	1	vect (100.0%)
VMOVUPD %ZMM5,-0x40(%R11)	2	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	1	1	5	2	vect (100.0%)
VMOVUPD (%R11),%ZMM6	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	1	vect (100.0%)
ADD $0x40,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %ZMM6,%ZMM6,%ZMM7	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	1	vect (100.0%)
VMOVUPD %ZMM7,-0x40(%R11)	2	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	1	1	5	2	vect (100.0%)
VMOVUPD (%R11),%ZMM8	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	1	vect (100.0%)
ADD $0x40,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %ZMM8,%ZMM8,%ZMM9	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	1	vect (100.0%)
VMOVUPD %ZMM9,-0x40(%R11)	2	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	1	1	5	2	vect (100.0%)
VMOVUPD (%R11),%ZMM10	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	1	vect (100.0%)
ADD $0x40,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %ZMM10,%ZMM10,%ZMM11	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	1	vect (100.0%)
VMOVUPD %ZMM11,-0x40(%R11)	2	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	1	1	5	2	vect (100.0%)
VMOVUPD (%R11),%ZMM12	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	1	vect (100.0%)
ADD $0x40,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %ZMM12,%ZMM12,%ZMM13	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	1	vect (100.0%)
VMOVUPD %ZMM13,-0x40(%R11)	2	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	1	1	5	2	vect (100.0%)
CMP %RCX,%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 4025d5 <main._omp_fn.3+0x215>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
MOV %RAX,%R10	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
AND $-0x8,%R10	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
ADD %R10,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
CMP %R10,%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JE 402660 <main._omp_fn.3+0x2a0>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
SUB %R10,%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
LEA -0x1(%RAX),%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
CMP $0x2,%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JBE 402615 <main._omp_fn.3+0x255>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
ADD %R9,%R10	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
LEA (%RDI,%R10,8),%R9	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
MOV %RAX,%R10	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
VMOVUPD (%R9),%YMM14	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	0.50	vect (50.0%)
AND $-0x4,%R10	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
ADD %R10,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VADDPD %YMM14,%YMM14,%YMM15	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	vect (50.0%)
VMOVUPD %YMM15,(%R9)	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0.50	0.50	4	1	vect (50.0%)
TEST $0x3,%AL	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
JE 402660 <main._omp_fn.3+0x2a0>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
LEA 0x1(%RDX),%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VMOVSD (%RDI,%RDX,8),%XMM0	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50	scal (12.5%)
VADDSD %XMM0,%XMM0,%XMM1	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	scal (12.5%)
VMOVSD %XMM1,(%RDI,%RDX,8)	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0.50	0.50	1	1	scal (12.5%)
CMP %RAX,%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JLE 402660 <main._omp_fn.3+0x2a0>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
ADD $0x2,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
VMOVSD (%RDI,%RAX,8),%XMM2	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50	scal (12.5%)
VADDSD %XMM2,%XMM2,%XMM3	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	scal (12.5%)
VMOVSD %XMM3,(%RDI,%RAX,8)	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0.50	0.50	1	1	scal (12.5%)
CMP %RDX,%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
JLE 402660 <main._omp_fn.3+0x2a0>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50-1	N/A
VMOVSD (%RDI,%RDX,8),%XMM4	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50	scal (12.5%)
VADDSD %XMM4,%XMM4,%XMM5	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0	3	0.50	scal (12.5%)
VMOVSD %XMM5,(%RDI,%RDX,8)	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0.50	0.50	1	1	scal (12.5%)
VZEROUPPER	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	vect (25.0%)
MOV -0x8(%RBP),%RBX	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	0.33	N/A
LEAVE	2	0.25	0.25	0.25	0.25	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50	N/A
RET	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50	N/A
NOPW (%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
VZEROUPPER	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	vect (25.0%)
MOV -0x8(%RBP),%RBX	1	0	0	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	3	0.33	N/A
LEAVE	2	0.25	0.25	0.25	0.25	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50	N/A
RET	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	1	0.50	N/A
NOPL (%RAX)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
INC %RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.25	N/A
XOR %EDX,%EDX	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.25	N/A
JMP 4023fa <main._omp_fn.3+0x3a>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	0	1	N/A
NOPW (%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
XOR %R10D,%R10D	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.25	N/A
MOV $0x334906100,%RDI	1	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
JMP 4025e4 <main._omp_fn.3+0x224>	1	0.50	0	0	0	0.50	0	0	0	0	0	0	0	0	0	0	1	N/A
NOPW %CS:(%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
NOPL (%RAX)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A

Name	Coverage (%)	Time (s)
▼main._omp_fn.3–	0.67	0.04
○Loop 3 - stream.c:292-292 - exec	0.67	0.04

Report Configuration

Code clean check

Workaround

Vectorization

Details

Workaround

Execution units bottlenecks

CALL instructions

Details

Complex instructions

Details

Vector unaligned load/store instructions

Details

Workaround

Conversion instructions

Details

Workaround

Type of elements and instruction set

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

Arithmetic intensity

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code

Code clean check

Workaround

Vectorization

Details

Workaround

Execution units bottlenecks

CALL instructions

Details

Complex instructions

Details

Vector unaligned load/store instructions

Details

Workaround

Conversion instructions

Details

Workaround

Type of elements and instruction set

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

Arithmetic intensity

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code