OV - exec - Function main.omp

	Function: main.omp_outlined.27	Module: exec	Source: stream.c:342-344	Coverage (incl. loops): 93.96% \| (excl. loops): 0.00%

/beegfs/hackathon/users/eoseret/qaas_runs_test/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/build/stream-add/src/stream.c: 342 - 344

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

342: #pragma omp parallel for

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

344: 	    c[j] = a[j]+b[j];

0x2034b0 PUSH	%RBP

0x2034b1 MOV	%RSP,%RBP

0x2034b4 PUSH	%R15

0x2034b6 PUSH	%R14

0x2034b8 PUSH	%R12

0x2034ba PUSH	%RBX

0x2034bb SUB	$0x20,%RSP

0x2034bf MOV	(%RDI),%R14D

0x2034c2 LEA	0x3067(%RIP),%RBX

0x2034c9 MOVQ	$0,-0x38(%RBP)

0x2034d1 MOVQ	$0x3344ffff,-0x28(%RBP)

0x2034d9 MOVQ	$0x1,-0x40(%RBP)

0x2034e1 MOVL	$0,-0x2c(%RBP)

0x2034e8 SUB	$0x8,%RSP

0x2034ec MOV	$0x1148,%RDI

0x2034f6 LEA	-0x40(%RBP),%RAX

0x2034fa LEA	-0x2c(%RBP),%RCX

0x2034fe LEA	-0x38(%RBP),%R8

0x203502 LEA	-0x28(%RBP),%R9

0x203506 MOV	%R14D,%ESI

0x203509 MOV	$0x22,%EDX

0x20350e ADD	%RBX,%RDI

0x203511 PUSH	$0x1

0x203513 PUSH	$0x1

0x203515 PUSH	%RAX

0x203516 CALL	204330 <@plt_start@+0x80>

0x20351b ADD	$0x20,%RSP

0x20351f MOV	-0x28(%RBP),%RCX

0x203523 MOV	-0x38(%RBP),%R15

0x203527 MOV	$0x3344ffff,%EAX

0x20352c CMP	$0x3344ffff,%RCX

0x203533 CMOVL	%RCX,%RAX

0x203537 MOV	%RAX,-0x28(%RBP)

0x20353b CMP	%RAX,%R15

0x20353e JG	20363a

0x203544 MOV	%RAX,%R8

0x203547 SUB	%R15,%R8

0x20354a MOV	$0x1190,%RCX

0x203554 MOV	$0x19a281190,%RDX

0x20355e MOV	$0x334501190,%RSI

0x203568 INC	%R8

0x20356b CMP	$0x8,%R8

0x20356f JAE	203579

0x203571 MOV	%R15,%RDI

0x203574 JMP	20360f

0x203579 MOV	%R8,%R9

0x20357c LEA	(%RBX,%RCX,1),%R10

0x203580 LEA	(%RBX,%RDX,1),%R11

0x203584 AND	$-0x8,%R9

0x203588 LEA	(%RBX,%RSI,1),%R12

0x20358c LEA	(%R15,%R9,1),%RDI

0x203590 LEA	0x30(%R10,%R15,8),%R10

0x203595 LEA	0x30(%R11,%R15,8),%R11

0x20359a LEA	0x30(%R12,%R15,8),%R15

0x20359f XOR	%R12D,%R12D

0x2035a2 NOPW	%CS:(%RAX,%RAX,1)

(8) 0x2035b0 VMOVUPD	-0x30(%R11,%R12,8),%XMM0

(8) 0x2035b7 VMOVUPD	-0x20(%R11,%R12,8),%XMM1

(8) 0x2035be VMOVUPD	-0x10(%R11,%R12,8),%XMM2

(8) 0x2035c5 VMOVUPD	(%R11,%R12,8),%XMM3

(8) 0x2035cb VADDPD	-0x30(%R10,%R12,8),%XMM0,%XMM0

(8) 0x2035d2 VADDPD	-0x20(%R10,%R12,8),%XMM1,%XMM1

(8) 0x2035d9 VADDPD	-0x10(%R10,%R12,8),%XMM2,%XMM2

(8) 0x2035e0 VADDPD	(%R10,%R12,8),%XMM3,%XMM3

(8) 0x2035e6 VMOVUPD	%XMM0,-0x30(%R15,%R12,8)

(8) 0x2035ed VMOVUPD	%XMM1,-0x20(%R15,%R12,8)

(8) 0x2035f4 VMOVUPD	%XMM2,-0x10(%R15,%R12,8)

(8) 0x2035fb VMOVUPD	%XMM3,(%R15,%R12,8)

(8) 0x203601 ADD	$0x8,%R12

(8) 0x203605 CMP	%R12,%R9

(8) 0x203608 JNE	2035b0

0x20360a CMP	%R9,%R8

0x20360d JE	20363a

0x20360f DEC	%RDI

0x203612 ADD	%RBX,%RCX

0x203615 ADD	%RBX,%RDX

0x203618 ADD	%RBX,%RSI

0x20361b NOPL	(%RAX,%RAX,1)

(9) 0x203620 VMOVSD	0x8(%RDX,%RDI,8),%XMM0

(9) 0x203626 VADDSD	0x8(%RCX,%RDI,8),%XMM0,%XMM0

(9) 0x20362c VMOVSD	%XMM0,0x8(%RSI,%RDI,8)

(9) 0x203632 INC	%RDI

(9) 0x203635 CMP	%RAX,%RDI

(9) 0x203638 JL	203620

0x20363a MOV	$0x1160,%RAX

0x203644 MOV	%R14D,%ESI

0x203647 ADD	%RAX,%RBX

0x20364a MOV	%RBX,%RDI

0x20364d CALL	204350 <@plt_start@+0xa0>

0x203652 ADD	$0x20,%RSP

0x203656 POP	%RBX

0x203657 POP	%R12

0x203659 POP	%R14

0x20365b POP	%R15

0x20365d POP	%RBP

0x20365e RET

0x20365f NOP

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
►99.67+	__kmp_invoke_microtask		libomp.so
○	__kmp_invoke_task_func		libomp.so

Coverage (%)	Name	Source Location	Module
►99.74+	__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:342-344
Module	exec

The function is defined in /beegfs/hackathon/users/eoseret/qaas_runs_test/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/build/stream-add/src/stream.c:342-344.

Warnings:
Ignoring paths for analysis

gain
potential
hint
expert

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 13.00 to 0.89 cycles (14.65x 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/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/build/stream-add/src/stream.c:342
- /beegfs/hackathon/users/eoseret/qaas_runs_test/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/build/stream-add/src/stream.c:342

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.

LEA: 3 occurrences►
- /beegfs/hackathon/users/eoseret/qaas_runs_test/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/build/stream-add/src/stream.c:342
- /beegfs/hackathon/users/eoseret/qaas_runs_test/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/build/stream-add/src/stream.c:342
- /beegfs/hackathon/users/eoseret/qaas_runs_test/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/build/stream-add/src/stream.c:342

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 20 bytes. The binary function is storing 36 bytes.

General properties

nb instructions	76
nb uops	78
loop length	316
used x86 registers	15
used mmx registers	0
used xmm registers	0
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	13.00 cycles
front end	13.00 cycles

Back-end

	ALU0/BRU0	ALU1	ALU2	ALU3	BRU1	AGU0	AGU1	AGU2	FP0	FP1	FP2	FP3	FP4	FP5
uops	10.25	10.25	10.00	10.00	3.50	4.33	4.33	4.33	0.00	0.00	0.00	0.00	0.00	0.00
cycles	10.25	10.25	10.00	10.00	3.50	4.33	4.33	4.33	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	13.00
Dispatch	10.25
Overall L1	13.00

Vectorization ratios

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

all	10%
load	6%
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	12%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 13.00 cycles. At this rate:

1% of peak load performance is reached (1.54 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
4% of peak store performance is reached (2.77 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 13.00 to 10.25 cycles (1.27x speedup).

ASM code

In the binary file, the address of the function is: 2034b0

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 %R12	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 $0x20,%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 0x3067(%RIP),%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
MOVQ $0,-0x38(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (6.3%)
MOVQ $0x3344ffff,-0x28(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (6.3%)
MOVQ $0x1,-0x40(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (6.3%)
MOVL $0,-0x2c(%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 -0x40(%RBP),%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA -0x2c(%RBP),%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA -0x38(%RBP),%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA -0x28(%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	N/A
MOV $0x22,%EDX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
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 204330 <@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 -0x28(%RBP),%RCX	1	0	0	0	0	0	0.33	0.33	0.33	3	0.33	N/A
MOV -0x38(%RBP),%R15	1	0	0	0	0	0	0.33	0.33	0.33	3	0.33	N/A
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 %RAX,-0x28(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (12.5%)
CMP %RAX,%R15	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
JG 20363a <main.omp_outlined.27+0x18a>	1	0.50	0	0	0	0.50	0	0	0	1	0.50-1	N/A
MOV %RAX,%R8	1	0	0	0	0	0	0	0	0	0	0.17	scal (12.5%)
SUB %R15,%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
MOV $0x1190,%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.33	N/A
MOV $0x19a281190,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.33	N/A
MOV $0x334501190,%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.33	N/A
INC %R8	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
CMP $0x8,%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
JAE 203579 <main.omp_outlined.27+0xc9>	1	0.50	0	0	0	0.50	0	0	0	1	0.50-1	N/A
MOV %R15,%RDI	1	0	0	0	0	0	0	0	0	0	0.17	N/A
JMP 20360f <main.omp_outlined.27+0x15f>	1	0.50	0	0	0	0.50	0	0	0	0	1	N/A
MOV %R8,%R9	1	0	0	0	0	0	0	0	0	0	0.17	N/A
LEA (%RBX,%RCX,1),%R10	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA (%RBX,%RDX,1),%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
AND $-0x8,%R9	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA (%RBX,%RSI,1),%R12	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA (%R15,%R9,1),%RDI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA 0x30(%R10,%R15,8),%R10	2	0	0	0.50	0.50	0	0.33	0.33	0.33	2	0.50	N/A
LEA 0x30(%R11,%R15,8),%R11	2	0	0	0.50	0.50	0	0.33	0.33	0.33	2	0.50	N/A
LEA 0x30(%R12,%R15,8),%R15	2	0	0	0.50	0.50	0	0.33	0.33	0.33	2	0.50	N/A
XOR %R12D,%R12D	1	0	0	0	0	0	0	0	0	0	0.25	N/A
NOPW %CS:(%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0.09	N/A
CMP %R9,%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
JE 20363a <main.omp_outlined.27+0x18a>	1	0.50	0	0	0	0.50	0	0	0	1	0.50-1	N/A
DEC %RDI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
ADD %RBX,%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
ADD %RBX,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
ADD %RBX,%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
NOPL (%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	N/A
ADD %RAX,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
MOV %RBX,%RDI	1	0	0	0	0	0	0	0	0	0	0.17	N/A
CALL 204350 <@plt_start@+0xa0>	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
POP %RBX	1	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R12	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
NOP	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:342-344
Module	exec

gain
potential
hint
expert

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/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/build/stream-add/src/stream.c:342
- /beegfs/hackathon/users/eoseret/qaas_runs_test/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/build/stream-add/src/stream.c:342

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.

LEA: 3 occurrences►
- /beegfs/hackathon/users/eoseret/qaas_runs_test/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/build/stream-add/src/stream.c:342
- /beegfs/hackathon/users/eoseret/qaas_runs_test/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/build/stream-add/src/stream.c:342
- /beegfs/hackathon/users/eoseret/qaas_runs_test/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/build/stream-add/src/stream.c:342

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 20 bytes. The binary function is storing 36 bytes.

General properties

nb instructions	76
nb uops	78
loop length	316
used x86 registers	15
used mmx registers	0
used xmm registers	0
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	13.00 cycles
front end	13.00 cycles

Back-end

	ALU0/BRU0	ALU1	ALU2	ALU3	BRU1	AGU0	AGU1	AGU2	FP0	FP1	FP2	FP3	FP4	FP5
uops	10.25	10.25	10.00	10.00	3.50	4.33	4.33	4.33	0.00	0.00	0.00	0.00	0.00	0.00
cycles	10.25	10.25	10.00	10.00	3.50	4.33	4.33	4.33	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	13.00
Dispatch	10.25
Overall L1	13.00

Vectorization ratios

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

all	10%
load	6%
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	12%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 13.00 cycles. At this rate:

1% of peak load performance is reached (1.54 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
4% of peak store performance is reached (2.77 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: 2034b0

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 %R12	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 $0x20,%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 0x3067(%RIP),%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
MOVQ $0,-0x38(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (6.3%)
MOVQ $0x3344ffff,-0x28(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (6.3%)
MOVQ $0x1,-0x40(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (6.3%)
MOVL $0,-0x2c(%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 -0x40(%RBP),%RAX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA -0x2c(%RBP),%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA -0x38(%RBP),%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA -0x28(%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	N/A
MOV $0x22,%EDX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
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 204330 <@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 -0x28(%RBP),%RCX	1	0	0	0	0	0	0.33	0.33	0.33	3	0.33	N/A
MOV -0x38(%RBP),%R15	1	0	0	0	0	0	0.33	0.33	0.33	3	0.33	N/A
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 %RAX,-0x28(%RBP)	1	0	0	0	0	0	0.33	0.33	0.33	4	0.50	scal (12.5%)
CMP %RAX,%R15	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
JG 20363a <main.omp_outlined.27+0x18a>	1	0.50	0	0	0	0.50	0	0	0	1	0.50-1	N/A
MOV %RAX,%R8	1	0	0	0	0	0	0	0	0	0	0.17	scal (12.5%)
SUB %R15,%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
MOV $0x1190,%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.33	N/A
MOV $0x19a281190,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.33	N/A
MOV $0x334501190,%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.33	N/A
INC %R8	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
CMP $0x8,%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
JAE 203579 <main.omp_outlined.27+0xc9>	1	0.50	0	0	0	0.50	0	0	0	1	0.50-1	N/A
MOV %R15,%RDI	1	0	0	0	0	0	0	0	0	0	0.17	N/A
JMP 20360f <main.omp_outlined.27+0x15f>	1	0.50	0	0	0	0.50	0	0	0	0	1	N/A
MOV %R8,%R9	1	0	0	0	0	0	0	0	0	0	0.17	N/A
LEA (%RBX,%RCX,1),%R10	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA (%RBX,%RDX,1),%R11	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
AND $-0x8,%R9	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA (%RBX,%RSI,1),%R12	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA (%R15,%R9,1),%RDI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
LEA 0x30(%R10,%R15,8),%R10	2	0	0	0.50	0.50	0	0.33	0.33	0.33	2	0.50	N/A
LEA 0x30(%R11,%R15,8),%R11	2	0	0	0.50	0.50	0	0.33	0.33	0.33	2	0.50	N/A
LEA 0x30(%R12,%R15,8),%R15	2	0	0	0.50	0.50	0	0.33	0.33	0.33	2	0.50	N/A
XOR %R12D,%R12D	1	0	0	0	0	0	0	0	0	0	0.25	N/A
NOPW %CS:(%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0.09	N/A
CMP %R9,%R8	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	scal (12.5%)
JE 20363a <main.omp_outlined.27+0x18a>	1	0.50	0	0	0	0.50	0	0	0	1	0.50-1	N/A
DEC %RDI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
ADD %RBX,%RCX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
ADD %RBX,%RDX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
ADD %RBX,%RSI	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
NOPL (%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	N/A
ADD %RAX,%RBX	1	0.25	0.25	0.25	0.25	0	0	0	0	1	0.25	N/A
MOV %RBX,%RDI	1	0	0	0	0	0	0	0	0	0	0.17	N/A
CALL 204350 <@plt_start@+0xa0>	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
POP %RBX	1	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R12	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
NOP	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/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/run/oneview_runs/multicore/aocc_4/oneview_run_1773741893OMP_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/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/run/oneview_runs/multicore/aocc_4/oneview_run_1773741893OMP_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/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/run/oneview_runs/multicore/aocc_4/oneview_run_1773741893OMP_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/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/run/oneview_runs/multicore/aocc_4/oneview_run_1773741893OMP_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/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/run/oneview_runs/multicore/aocc_4/oneview_run_1773741893OMP_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/ins01.benchmarkcenter.megware.com/177-374-0840/stream-add/run/oneview_runs/multicore/aocc_4/oneview_run_1773741893OMP_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	67.87	0.15	80.89	0.12	79.46	0.09	81.86	0.08	88.15

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	6.9150004386902	93.956283569336
1x64	64	0.24	1.93	8	3.7400002479553	89.512634277344
1x96	96	0.15	1.85	12	3.6149997711182	95.682823181152
1x128	128	0.12	1.88	16	3.7699997425079	90.030052185059
1x160	160	0.09	1.86	20	3.7700004577637	90.23902130127
1x192	192	0.08	1.82	24	3.6099996566772	95.361167907715

Name	Coverage (%)	Time (s)
▼main.omp_outlined.27–	93.96	6.79
○Loop 8 - stream.c:344-344 - exec	93.96	6.74
○Loop 9 - stream.c:343-344 - exec	0.00	0.00

Report Configuration

Vectorization

Details

Workaround

Execution units bottlenecks

CALL instructions

Details

Complex 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

Vectorization

Details

Workaround

Execution units bottlenecks

CALL instructions

Details

Complex 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