OV - exec - Function main.omp

	Function: main.omp_outlined.27	Module: exec	Source: stream.c:330-332	Coverage (incl. loops): 93.55% \| (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: 330 - 332

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

330: #pragma omp parallel for

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

332: 	    b[j] = scalar*c[j];

0x203500 PUSH	%RBP

0x203501 MOV	%RSP,%RBP

0x203504 PUSH	%R15

0x203506 PUSH	%R14

0x203508 PUSH	%RBX

0x203509 SUB	$0x28,%RSP

0x20350d MOV	(%RDI),%R14D

0x203510 LEA	0x2cc9(%RIP),%RBX

0x203517 MOV	%RDX,%R15

0x20351a MOVQ	$0,-0x30(%RBP)

0x203522 MOVQ	$0x3344ffff,-0x20(%RBP)

0x20352a MOVQ	$0x1,-0x38(%RBP)

0x203532 MOVL	$0,-0x24(%RBP)

0x203539 SUB	$0x8,%RSP

0x20353d MOV	$0x1148,%RDI

0x203547 LEA	-0x38(%RBP),%RAX

0x20354b LEA	-0x24(%RBP),%RCX

0x20354f LEA	-0x30(%RBP),%R8

0x203553 LEA	-0x20(%RBP),%R9

0x203557 MOV	%R14D,%ESI

0x20355a MOV	$0x22,%EDX

0x20355f ADD	%RBX,%RDI

0x203562 PUSH	$0x1

0x203564 PUSH	$0x1

0x203566 PUSH	%RAX

0x203567 CALL	203fe0 <@plt_start@+0x80>

0x20356c ADD	$0x20,%RSP

0x203570 MOV	-0x20(%RBP),%RCX

0x203574 MOV	$0x3344ffff,%EAX

0x203579 CMP	$0x3344ffff,%RCX

0x203580 CMOVL	%RCX,%RAX

0x203584 MOV	-0x30(%RBP),%RCX

0x203588 MOV	%RAX,-0x20(%RBP)

0x20358c CMP	%RAX,%RCX

0x20358f JG	2035d4

0x203591 VMOVSD	(%R15),%XMM0

0x203596 MOV	$0x334501190,%RDX

0x2035a0 MOV	$0x19a281190,%RSI

0x2035aa DEC	%RCX

0x2035ad ADD	%RBX,%RDX

0x2035b0 ADD	%RBX,%RSI

0x2035b3 NOPW	%CS:(%RAX,%RAX,1)

(6) 0x2035c0 VMULSD	0x8(%RDX,%RCX,8),%XMM0,%XMM1

(6) 0x2035c6 VMOVSD	%XMM1,0x8(%RSI,%RCX,8)

(6) 0x2035cc INC	%RCX

(6) 0x2035cf CMP	%RAX,%RCX

(6) 0x2035d2 JL	2035c0

0x2035d4 MOV	$0x1160,%RAX

0x2035de MOV	%R14D,%ESI

0x2035e1 ADD	%RAX,%RBX

0x2035e4 MOV	%RBX,%RDI

0x2035e7 CALL	204000 <@plt_start@+0xa0>

0x2035ec ADD	$0x28,%RSP

0x2035f0 POP	%RBX

0x2035f1 POP	%R14

0x2035f3 POP	%R15

0x2035f5 POP	%RBP

0x2035f6 RET

0x2035f7 NOPW	(%RAX,%RAX,1)

Coverage (%)	Name	Source Location	Module
►100.00+	__kmp_invoke_microtask		libomp.so
○	__kmp_invoke_task_func		libomp.so

Coverage (%)	Name	Source Location	Module
►100.00+	__kmp_invoke_microtask		libomp.so
○	__kmp_invoke_task_func		libomp.so

Coverage (%)	Name	Source Location	Module
►100.00+	__kmp_invoke_microtask		libomp.so
○	__kmp_invoke_task_func		libomp.so

Coverage (%)	Name	Source Location	Module
►99.66+	__kmp_invoke_microtask		libomp.so
○	__kmp_invoke_task_func		libomp.so

Coverage (%)	Name	Source Location	Module
►99.73+	__kmp_invoke_microtask		libomp.so
○	__kmp_invoke_task_func		libomp.so

Coverage (%)	Name	Source Location	Module
►100.00+	__kmp_invoke_microtask		libomp.so
○	__kmp_invoke_task_func		libomp.so

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 1x8

Source file and lines	stream.c:330-332
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:330-332.

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 9.00 to 1.25 cycles (7.20x speedup).

Workaround

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

Vectorization

Your function is not vectorized. Only 10% of vector register length is used (average across all SSE/AVX instructions). By vectorizing your function, you can lower the cost of an iteration from 9.00 to 0.67 cycles (13.50x speedup).

Details

All SSE/AVX instructions are used in scalar version (process only one data element in vector registers). Since your execution units are vector units, only a 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:330
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:330

Type of elements and instruction set

No instructions are processing arithmetic or math operations on FP elements. This function is probably writing/copying data or processing integer elements.

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

The binary function does not contain any FP arithmetical operations. The binary function is loading 28 bytes. The binary function is storing 36 bytes.

General properties

nb instructions	54
nb uops	54
loop length	236
used x86 registers	12
used mmx registers	0
used xmm registers	1
used ymm registers	0
used zmm registers	0
nb stack references	4

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	9.00 cycles
front end	9.00 cycles

Back-end

	ALU0/BRU0	ALU1	ALU2	ALU3	BRU1	AGU0	AGU1	AGU2	FP0	FP1	FP2	FP3	FP4	FP5
uops	6.25	6.25	6.25	6.25	2.00	3.67	3.67	3.67	0.00	0.00	0.00	0.00	0.00	0.00
cycles	6.25	6.25	6.25	6.25	2.00	3.67	3.67	3.67	0.00	0.00	0.00	0.00	0.00	0.00

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

Cycles summary

Front-end	9.00
Dispatch	6.25
Overall L1	9.00

Vectorization ratios

INT

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

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

Vector efficiency ratios

INT

all	9%
load	10%
store	7%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	12%
fma	NA (no fma vectorizable/vectorized instructions)
other	9%

all	12%
load	12%
store	NA (no store vectorizable/vectorized instructions)
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	NA (no add-sub vectorizable/vectorized instructions)
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	10%
load	10%
store	7%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	12%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	9%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 9.00 cycles. At this rate:

3% of peak load performance is reached (3.11 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
6% of peak store performance is reached (4.00 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 9.00 to 6.25 cycles (1.44x speedup).

ASM code

In the binary file, the address of the function is: 203500

Instruction	Nb FU	ALU0/BRU0	ALU1	ALU2	ALU3	BRU1	AGU0	AGU1	AGU2	Latency	Recip. throughput	Vectorization
PUSH %RBP	1	0	0	0	0	0	0	0	0	1	0.50	N/A
MOV %RSP,%RBP	1	0	0	0	0	0	0	0	0	0	0.17	N/A
PUSH %R15	1	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R14	1	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RBX	1	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB $0x28,%RSP	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
MOV (%RDI),%R14D	1	0	0	0	0	0	0.33	0.33	0.33	3	0.33	scal (6.3%)
LEA 0x2cc9(%RIP),%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
MOV %RDX,%R15	1	0	0	0	0	0	0	0	0	0	0.17	N/A
MOVQ $0,-0x30(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (6.3%)
MOVQ $0x3344ffff,-0x20(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (6.3%)
MOVQ $0x1,-0x38(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (6.3%)
MOVL $0,-0x24(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (6.3%)
SUB $0x8,%RSP	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
MOV $0x1148,%RDI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.33	N/A
LEA -0x38(%RBP),%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA -0x24(%RBP),%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA -0x30(%RBP),%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA -0x20(%RBP),%R9	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
MOV %R14D,%ESI	1	0	0	0	0	0	0	0	0	0	0.17	scal (6.3%)
MOV $0x22,%EDX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (6.3%)
ADD %RBX,%RDI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
PUSH $0x1	1	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH $0x1	1	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RAX	1	0	0	0	0	0	0	0	0	1	0.50	N/A
CALL 203fe0 <@plt_start@+0x80>	2	0.50	0	0	0	0.50	0.33	0.33	0.33	0	0.50	N/A
ADD $0x20,%RSP	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
MOV -0x20(%RBP),%RCX	1	0	0	0	0	0	0.33	0.33	0.33	3	0.33	scal (12.5%)
MOV $0x3344ffff,%EAX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
CMP $0x3344ffff,%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
CMOVL %RCX,%RAX	1	0.50	0	0	0.50	0	0	0	0	1	0.50	N/A
MOV -0x30(%RBP),%RCX	1	0	0	0	0	0	0.33	0.33	0.33	3	0.33	scal (12.5%)
MOV %RAX,-0x20(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (12.5%)
CMP %RAX,%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
JG 2035d4 <main.omp_outlined.27+0xd4>	1	0.50	0	0	0	0.50	0	0	0	1	0.50-1	N/A
VMOVSD (%R15),%XMM0	1	0	0	0	0	0	0.33	0.33	0.33	1	0.50	scal (12.5%)
MOV $0x334501190,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.33	scal (12.5%)
MOV $0x19a281190,%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.33	scal (12.5%)
DEC %RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
ADD %RBX,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
ADD %RBX,%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
NOPW %CS:(%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0.09	N/A
MOV $0x1160,%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.33	N/A
MOV %R14D,%ESI	1	0	0	0	0	0	0	0	0	0	0.17	scal (6.3%)
ADD %RAX,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
MOV %RBX,%RDI	1	0	0	0	0	0	0	0	0	0	0.17	N/A
CALL 204000 <@plt_start@+0xa0>	2	0.50	0	0	0	0.50	0.33	0.33	0.33	0	0.50	N/A
ADD $0x28,%RSP	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
POP %RBX	1	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R14	1	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R15	1	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %RBP	1	0	0	0	0	0	0	0	0	1	0.33	N/A
RET	1	0.50	0	0	0	0.50	0	0	0	1	0.50	N/A
NOPW (%RAX,%RAX,1)	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 1x8

Source file and lines	stream.c:330-332
Module	exec

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 9.00 to 1.25 cycles (7.20x speedup).

Workaround

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

Vectorization

Details

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:330
- /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/build/stream-scale/src/stream.c:330

Type of elements and instruction set

No instructions are processing arithmetic or math operations on FP elements. This function is probably writing/copying data or processing integer elements.

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

The binary function does not contain any FP arithmetical operations. The binary function is loading 28 bytes. The binary function is storing 36 bytes.

General properties

nb instructions	54
nb uops	54
loop length	236
used x86 registers	12
used mmx registers	0
used xmm registers	1
used ymm registers	0
used zmm registers	0
nb stack references	4

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	9.00 cycles
front end	9.00 cycles

Back-end

	ALU0/BRU0	ALU1	ALU2	ALU3	BRU1	AGU0	AGU1	AGU2	FP0	FP1	FP2	FP3	FP4	FP5
uops	6.25	6.25	6.25	6.25	2.00	3.67	3.67	3.67	0.00	0.00	0.00	0.00	0.00	0.00
cycles	6.25	6.25	6.25	6.25	2.00	3.67	3.67	3.67	0.00	0.00	0.00	0.00	0.00	0.00

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

Cycles summary

Front-end	9.00
Dispatch	6.25
Overall L1	9.00

Vectorization ratios

INT

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

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

Vector efficiency ratios

INT

all	9%
load	10%
store	7%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	12%
fma	NA (no fma vectorizable/vectorized instructions)
other	9%

all	12%
load	12%
store	NA (no store vectorizable/vectorized instructions)
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	NA (no add-sub vectorizable/vectorized instructions)
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	10%
load	10%
store	7%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	12%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	9%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 9.00 cycles. At this rate:

3% of peak load performance is reached (3.11 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
6% of peak store performance is reached (4.00 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: 203500

Instruction	Nb FU	ALU0/BRU0	ALU1	ALU2	ALU3	BRU1	AGU0	AGU1	AGU2	Latency	Recip. throughput	Vectorization
PUSH %RBP	1	0	0	0	0	0	0	0	0	1	0.50	N/A
MOV %RSP,%RBP	1	0	0	0	0	0	0	0	0	0	0.17	N/A
PUSH %R15	1	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R14	1	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RBX	1	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB $0x28,%RSP	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
MOV (%RDI),%R14D	1	0	0	0	0	0	0.33	0.33	0.33	3	0.33	scal (6.3%)
LEA 0x2cc9(%RIP),%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
MOV %RDX,%R15	1	0	0	0	0	0	0	0	0	0	0.17	N/A
MOVQ $0,-0x30(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (6.3%)
MOVQ $0x3344ffff,-0x20(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (6.3%)
MOVQ $0x1,-0x38(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (6.3%)
MOVL $0,-0x24(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (6.3%)
SUB $0x8,%RSP	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
MOV $0x1148,%RDI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.33	N/A
LEA -0x38(%RBP),%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA -0x24(%RBP),%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA -0x30(%RBP),%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA -0x20(%RBP),%R9	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
MOV %R14D,%ESI	1	0	0	0	0	0	0	0	0	0	0.17	scal (6.3%)
MOV $0x22,%EDX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (6.3%)
ADD %RBX,%RDI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
PUSH $0x1	1	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH $0x1	1	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RAX	1	0	0	0	0	0	0	0	0	1	0.50	N/A
CALL 203fe0 <@plt_start@+0x80>	2	0.50	0	0	0	0.50	0.33	0.33	0.33	0	0.50	N/A
ADD $0x20,%RSP	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
MOV -0x20(%RBP),%RCX	1	0	0	0	0	0	0.33	0.33	0.33	3	0.33	scal (12.5%)
MOV $0x3344ffff,%EAX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
CMP $0x3344ffff,%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
CMOVL %RCX,%RAX	1	0.50	0	0	0.50	0	0	0	0	1	0.50	N/A
MOV -0x30(%RBP),%RCX	1	0	0	0	0	0	0.33	0.33	0.33	3	0.33	scal (12.5%)
MOV %RAX,-0x20(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (12.5%)
CMP %RAX,%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
JG 2035d4 <main.omp_outlined.27+0xd4>	1	0.50	0	0	0	0.50	0	0	0	1	0.50-1	N/A
VMOVSD (%R15),%XMM0	1	0	0	0	0	0	0.33	0.33	0.33	1	0.50	scal (12.5%)
MOV $0x334501190,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.33	scal (12.5%)
MOV $0x19a281190,%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.33	scal (12.5%)
DEC %RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
ADD %RBX,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
ADD %RBX,%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
NOPW %CS:(%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0.09	N/A
MOV $0x1160,%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.33	N/A
MOV %R14D,%ESI	1	0	0	0	0	0	0	0	0	0	0.17	scal (6.3%)
ADD %RAX,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
MOV %RBX,%RDI	1	0	0	0	0	0	0	0	0	0	0.17	N/A
CALL 204000 <@plt_start@+0xa0>	2	0.50	0	0	0	0.50	0.33	0.33	0.33	0	0.50	N/A
ADD $0x28,%RSP	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
POP %RBX	1	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R14	1	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R15	1	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %RBP	1	0	0	0	0	0	0	0	0	1	0.33	N/A
RET	1	0.50	0	0	0	0.50	0	0	0	1	0.50	N/A
NOPW (%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0.09	N/A

▶Scalability Runs Description

Run 1x8	Number processes: 1Number nodes: NANumber processes per node: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to none --report-bindings Dataset: Run Directory: /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/run/oneview_runs/multicore/aocc_5/oneview_run_1773740879OMP_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: threadsOMP_NUM_THREADS: 8
Run 1x64	Number processes: 1Number processes per node: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to none --report-bindings Dataset: Run Directory: /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/run/oneview_runs/multicore/aocc_5/oneview_run_1773740879OMP_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
Run 1x96	Number processes: 1Number processes per node: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to none --report-bindings Dataset: Run Directory: /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/run/oneview_runs/multicore/aocc_5/oneview_run_1773740879OMP_NUM_THREADS: 96OMP_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 1x128	Number processes: 1Number processes per node: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to none --report-bindings Dataset: Run Directory: /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/run/oneview_runs/multicore/aocc_5/oneview_run_1773740879OMP_NUM_THREADS: 128OMP_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 1x160	Number processes: 1Number processes per node: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to none --report-bindings Dataset: Run Directory: /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/run/oneview_runs/multicore/aocc_5/oneview_run_1773740879OMP_NUM_THREADS: 160OMP_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 1x192	Number processes: 1Number processes per node: 1Run Command: <executable>MPI Command: mpirun -n <number_processes> --bind-to none --report-bindings Dataset: Run Directory: /beegfs/hackathon/users/eoseret/qaas_runs_test/gmz16.benchmarkcenter.megware.com/177-374-0012/stream-scale/run/oneview_runs/multicore/aocc_5/oneview_run_1773740879OMP_NUM_THREADS: 192OMP_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

(1x8) Efficiency	(1x8) Potential Speed-Up (%)	(1x64) Efficiency	(1x64) Potential Speed-Up (%)	(1x96) Efficiency	(1x96) Potential Speed-Up (%)	(1x128) Efficiency	(1x128) Potential Speed-Up (%)	(1x160) Efficiency	(1x160) Potential Speed-Up (%)	(1x192) Efficiency	(1x192) Potential Speed-Up (%)
1	0	0.24	68.14	0.15	79.76	0.12	78.99	0.09	81.07	0.07	87.27

Speed-Up and Efficiency

▶Detailed Speed-Up and Efficiency

Run	Number of threads	Efficiency (ideal is 1)	Speedup	Ideal Speedup	Time (s)	Coverage (%)
1x8	8	1	1	1	5.2800002098083	93.550552368164
1x64	64	0.24	1.93	8	2.8450000286102	89.837188720703
1x96	96	0.15	1.85	12	2.7950000762939	94.337844848633
1x128	128	0.12	1.87	16	2.8899998664856	89.460014343262
1x160	160	0.09	1.85	20	2.9100003242493	89.315994262695
1x192	192	0.07	1.8	24	2.7950003147125	94.326858520508

Name	Coverage (%)	Time (s)
▼main.omp_outlined.27–	93.55	5.27
○Loop 6 - stream.c:331-332 - exec	93.55	5.22

Report Configuration

Code clean check

Workaround

Vectorization

Details

Workaround

Execution units bottlenecks

CALL instructions

Details

Type of elements and instruction set

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

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

Type of elements and instruction set

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

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