OV - exec - Function qmcplusplus::TimerType<std::chrono::_V2::system

	Function: qmcplusplus::TimerType<std::chrono::_V2::system_clock>::start()	Module: exec	Source: NewTimer.cpp:39-81 [...]	Coverage (incl. loops): 0.01% \| (excl. loops): 0.01%

/home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/TimerManager.h: 92 - 93

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

92:     if (CurrentTimerStack.size() > 0)

93:       current = CurrentTimerStack.back();

/home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp: 39 - 81

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

39:   if (active)

[...]

53:     for (int level = omp_get_level(); level > 0; level--)

54:       if (omp_get_ancestor_thread_num(level) != 0)

55:         is_true_master = false;

56:     if (is_true_master)

57:     {

58:       if (manager)

59:       {

60:         // compute current_stack_key

61:         TimerType* parent = manager->current_timer();

62:         if (parent)

63:         {

64:           current_stack_key = parent->get_stack_key();

65:           current_stack_key.add_id(timer_id);

66:         }

67:         else

68:         {

69:           current_stack_key = StackKey();

70:           current_stack_key.add_id(timer_id);

71:         }

72:

73:         manager->push_timer(this);

74:       }

75:       start_time = CLOCK::now();

[...]

81: }

/home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.h: 90 - 97

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

90:     short_buckets[level] = c1;

91:     if (level >= max_level - 1)

92:     {

93:       timer_max_level_exceeded = true;

94:     }

95:     else

96:     {

97:       level++;

/usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_vector.h: 919 - 919

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

919:       { return size_type(this->_M_impl._M_finish - this->_M_impl._M_start); }

0x4a9a00 CMPB	$0,0x38(%RDI)

0x4a9a04 JE	4a9bb8

0x4a9a0a PUSH	%RBP

0x4a9a0b MOV	%RSP,%RBP

0x4a9a0e PUSH	%R15

0x4a9a10 PUSH	%R14

0x4a9a12 PUSH	%R13

0x4a9a14 PUSH	%R12

0x4a9a16 PUSH	%RBX

0x4a9a17 SUB	$0x18,%RSP

0x4a9a1b MOV	%RDI,%R15

0x4a9a1e CALL	405970 <omp_get_level@plt>

0x4a9a23 TEST	%EAX,%EAX

0x4a9a25 JLE	4a9b2c

0x4a9a2b MOV	%R15,-0x38(%RBP)

0x4a9a2f MOV	$0x1,%R14B

0x4a9a32 MOV	%EAX,-0x2c(%RBP)

0x4a9a35 CMP	$0x8,%EAX

0x4a9a38 JB	4a9add

0x4a9a3e MOV	-0x2c(%RBP),%EAX

0x4a9a41 MOV	%EAX,%ECX

0x4a9a43 SHR	$0x3,%ECX

0x4a9a46 MOV	$0x1,%R14B

0x4a9a49 MOV	%EAX,%R15D

0x4a9a4c NOPL	(%RAX)

(1703) 0x4a9a50 MOV	%ECX,-0x30(%RBP)

(1703) 0x4a9a53 MOV	%R15D,%EDI

(1703) 0x4a9a56 CALL	405360 <omp_get_ancestor_thread_num@plt>

(1703) 0x4a9a5b MOV	%EAX,%R12D

(1703) 0x4a9a5e LEA	-0x1(%R15),%EDI

(1703) 0x4a9a62 CALL	405360 <omp_get_ancestor_thread_num@plt>

(1703) 0x4a9a67 MOV	%EAX,%R13D

(1703) 0x4a9a6a OR	%R12D,%R13D

(1703) 0x4a9a6d LEA	-0x2(%R15),%EDI

(1703) 0x4a9a71 CALL	405360 <omp_get_ancestor_thread_num@plt>

(1703) 0x4a9a76 MOV	%EAX,%R12D

(1703) 0x4a9a79 LEA	-0x3(%R15),%EDI

(1703) 0x4a9a7d CALL	405360 <omp_get_ancestor_thread_num@plt>

(1703) 0x4a9a82 MOV	%EAX,%EBX

(1703) 0x4a9a84 OR	%R12D,%EBX

(1703) 0x4a9a87 OR	%R13D,%EBX

(1703) 0x4a9a8a LEA	-0x4(%R15),%EDI

(1703) 0x4a9a8e CALL	405360 <omp_get_ancestor_thread_num@plt>

(1703) 0x4a9a93 MOV	%EAX,%R12D

(1703) 0x4a9a96 LEA	-0x5(%R15),%EDI

(1703) 0x4a9a9a CALL	405360 <omp_get_ancestor_thread_num@plt>

(1703) 0x4a9a9f MOV	%EAX,%R13D

(1703) 0x4a9aa2 OR	%R12D,%R13D

(1703) 0x4a9aa5 OR	%EBX,%R13D

(1703) 0x4a9aa8 LEA	-0x6(%R15),%EDI

(1703) 0x4a9aac CALL	405360 <omp_get_ancestor_thread_num@plt>

(1703) 0x4a9ab1 MOV	%EAX,%EBX

(1703) 0x4a9ab3 LEA	-0x7(%R15),%EDI

(1703) 0x4a9ab7 CALL	405360 <omp_get_ancestor_thread_num@plt>

(1703) 0x4a9abc MOV	-0x30(%RBP),%ECX

(1703) 0x4a9abf OR	%EBX,%EAX

(1703) 0x4a9ac1 MOVZX	%R14B,%R14D

(1703) 0x4a9ac5 OR	%R13D,%EAX

(1703) 0x4a9ac8 MOV	$0,%EAX

(1703) 0x4a9acd CMOVNE	%EAX,%R14D

(1703) 0x4a9ad1 ADD	$-0x8,%R15D

(1703) 0x4a9ad5 DEC	%ECX

(1703) 0x4a9ad7 JNE	4a9a50

0x4a9add MOV	-0x2c(%RBP),%R12D

0x4a9ae1 MOV	%R12D,%EAX

0x4a9ae4 AND	$-0x8,%EAX

0x4a9ae7 SUB	%EAX,%R12D

0x4a9aea JBE	4a9b22

0x4a9aec XOR	%EBX,%EBX

0x4a9aee MOV	-0x38(%RBP),%R15

0x4a9af2 NOPW	%CS:(%RAX,%RAX,1)

(1702) 0x4a9b00 MOV	%R12D,%EDI

(1702) 0x4a9b03 CALL	405360 <omp_get_ancestor_thread_num@plt>

(1702) 0x4a9b08 TEST	%EAX,%EAX

(1702) 0x4a9b0a MOVZX	%R14B,%R14D

(1702) 0x4a9b0e CMOVNE	%EBX,%R14D

(1702) 0x4a9b12 DEC	%R12D

(1702) 0x4a9b15 JNE	4a9b00

0x4a9b17 TEST	$0x1,%R14B

0x4a9b1b JNE	4a9b2c

0x4a9b1d JMP	4a9baa

0x4a9b22 MOV	-0x38(%RBP),%R15

0x4a9b26 TEST	$0x1,%R14B

0x4a9b2a JE	4a9baa

0x4a9b2c MOV	0x48(%R15),%RAX

0x4a9b30 TEST	%RAX,%RAX

0x4a9b33 JE	4a9ba2

0x4a9b35 MOV	0x48(%RAX),%RCX

0x4a9b39 CMP	0x40(%RAX),%RCX

0x4a9b3d JE	4a9b75

0x4a9b3f MOV	-0x8(%RCX),%RAX

0x4a9b43 TEST	%RAX,%RAX

0x4a9b46 JE	4a9b75

0x4a9b48 MOVUPS	0x50(%RAX),%XMM0

0x4a9b4c MOVUPS	%XMM0,0x50(%R15)

0x4a9b51 MOVSXD	0x60(%RAX),%RAX

0x4a9b55 MOV	%EAX,0x60(%R15)

0x4a9b59 MOVZX	0x40(%R15),%ECX

0x4a9b5e MOV	%CL,0x50(%R15,%RAX,1)

0x4a9b63 MOV	0x60(%R15),%EAX

0x4a9b67 CMP	$0xf,%EAX

0x4a9b6a JL	4a9b90

0x4a9b6c MOVB	$0x1,0x102864(%RIP)

0x4a9b73 JMP	4a9b96

0x4a9b75 XORPS	%XMM0,%XMM0

0x4a9b78 MOVUPS	%XMM0,0x50(%R15)

0x4a9b7d MOVZX	0x40(%R15),%EAX

0x4a9b82 MOV	%AL,0x50(%R15)

0x4a9b86 MOVL	$0x1,0x60(%R15)

0x4a9b8e JMP	4a9b96

0x4a9b90 INC	%EAX

0x4a9b92 MOV	%EAX,0x60(%R15)

0x4a9b96 MOV	0x48(%R15),%RDI

0x4a9b9a MOV	%R15,%RSI

0x4a9b9d CALL	4af8d0 <_ZN11qmcplusplus12TimerManagerINS_9TimerTypeINSt6chrono3_V212system_clockEEEE10push_timerEPS5_>

0x4a9ba2 CALL	405a60 <_ZNSt6chrono3_V212system_clock3nowEv@plt>

0x4a9ba7 MOV	%RAX,(%R15)

0x4a9baa ADD	$0x18,%RSP

0x4a9bae POP	%RBX

0x4a9baf POP	%R12

0x4a9bb1 POP	%R13

0x4a9bb3 POP	%R14

0x4a9bb5 POP	%R15

0x4a9bb7 POP	%RBP

0x4a9bb8 RET

0x4a9bb9 NOPL	(%RAX)

Coverage (%)	Name	Source Location	Module
►13.46+	_ZN11qmcplusplus11ParticleSet6[...]	unique_ptr.h:173	exec
○	qmcplusplus::NonLocalPP<double[...]	NonLocalPP.hpp:134	exec
○	main.extracted.107	NewTimer.h:249	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►12.20+	qmcplusplus::SoaDistanceTableA[...]	SoaDistanceTableABOMPTarget.h:190	exec
○	_ZN11qmcplusplus11ParticleSet6[...]	ParticleSet.cpp:242	exec
○	qmcplusplus::NonLocalPP<double[...]	NonLocalPP.hpp:134	exec
○	main.extracted.107	NewTimer.h:249	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►9.36+	_ZN11qmcplusplus12WaveFunction[...]	stl_vector.h:1046	exec
○	main.extracted.107	refwrap.h:338	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►8.31+	qmcplusplus::SoaDistanceTableA[...]	SoaDistanceTableABOMPTarget.h:374	exec
○	_ZN11qmcplusplus11ParticleSet1[...]	ParticleSet.cpp:389	exec
○	main.extracted.107	miniqmc.cpp:452	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►7.26+	_ZN11qmcplusplus12WaveFunction[...]	stl_vector.h:1046	exec
○	qmcplusplus::NonLocalPP<double[...]	NonLocalPP.hpp:126	exec
○	main.extracted.107	NewTimer.h:249	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►5.15+	_ZN11qmcplusplus12WaveFunction[...]	stl_vector.h:1046	exec
○	main.extracted.107	refwrap.h:338	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►5.15+	_ZN11qmcplusplus11ParticleSet1[...]	unique_ptr.h:173	exec
○	main.extracted.107	miniqmc.cpp:452	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►4.84+	miniqmcreference::einspline_sp[...]	ParticleSet.h:143	exec
○	qmcplusplus::SPOSet::evaluateD[...]	OhmmsVector.h:178	exec
○	miniqmcreference::DiracDetermi[...]	DiracDeterminantRef.cpp:217	exec
○	_ZN11qmcplusplus12WaveFunction[...]	stl_vector.h:919	exec
○	qmcplusplus::NonLocalPP<double[...]	NonLocalPP.hpp:126	exec
○	main.extracted.107	NewTimer.h:249	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►4.73+	main.extracted.107	miniqmc.cpp:433	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►3.26+	_ZN11qmcplusplus12WaveFunction[...]	stl_vector.h:1046	exec
○	main.extracted.107	refwrap.h:338	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►2.73+	qmcplusplus::SoaDistanceTableA[...]	SoaDistanceTableAAOMPTarget.h:436	exec
○	_ZN11qmcplusplus11ParticleSet1[...]	ParticleSet.cpp:389	exec
○	main.extracted.107	miniqmc.cpp:452	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►2.42+	qmcplusplus::SoaDistanceTableA[...]	NewTimer.h:242	exec
○	_ZN11qmcplusplus11ParticleSet6[...]	ParticleSet.cpp:242	exec
○	qmcplusplus::NonLocalPP<double[...]	NonLocalPP.hpp:134	exec
○	main.extracted.107	NewTimer.h:249	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►2.21+	miniqmcreference::einspline_sp[...]	ParticleSet.h:143	exec
○	miniqmcreference::DiracDetermi[...]	DiracDeterminantRef.cpp:100	exec
○	_ZN11qmcplusplus12WaveFunction[...]	WaveFunction.cpp:196	exec
○	main.extracted.107	refwrap.h:338	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►1.89+	main.extracted.107	TinyVector.h:62	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►1.79+	_ZN11qmcplusplus11ParticleSet2[...]	stl_vector.h:919	exec
○	main.extracted.107	refwrap.h:338	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►1.58+	miniqmcreference::DiracDetermi[...]	DiracDeterminantRef.cpp:86	exec
○	_ZN11qmcplusplus12WaveFunction[...]	stl_vector.h:919	exec
○	main.extracted.107	refwrap.h:338	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►1.26+	miniqmcreference::DiracDetermi[...]	unique_ptr.h:173	exec
○	_ZN11qmcplusplus12WaveFunction[...]	WaveFunction.cpp:196	exec
○	main.extracted.107	refwrap.h:338	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►1.26+	miniqmcreference::DiracDetermi[...]	DiracDeterminantRef.cpp:213	exec
○	_ZN11qmcplusplus12WaveFunction[...]	stl_vector.h:919	exec
○	qmcplusplus::NonLocalPP<double[...]	NonLocalPP.hpp:126	exec
○	main.extracted.107	NewTimer.h:249	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►1.16+	qmcplusplus::SoaDistanceTableA[...]	OhmmsVector.h:178	exec
○	_ZN11qmcplusplus11ParticleSet2[...]	ParticleSet.cpp:343	exec
○	main.extracted.107	refwrap.h:338	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►1.05+	miniqmcreference::DiracDetermi[...]	DiracDeterminantRef.cpp:103	exec
○	_ZN11qmcplusplus12WaveFunction[...]	WaveFunction.cpp:196	exec
○	main.extracted.107	refwrap.h:338	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►1.05+	main.extracted.107	miniqmc.cpp:449	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
○	__kmp_launch_thread		libiomp5.so
○	_INTERNALea951ae6::__kmp_launc[...]		libiomp5.so
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6

min	med	avg	max

Percentile Index	10	20	30	40	50	60	70	80	90	100
Value

min	med	avg	max

Percentile Index	10	20	30	40	50	60	70	80	90	100
Value

Path /

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 0.01% of application time for run icx_3

Source file and lines	NewTimer.cpp:39-81
Module	exec

The function is defined in:

/home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/TimerManager.h:92-93
/home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:39,53-75,81
/home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.h:90-97
/usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_vector.h:919

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 10.13 to 3.33 cycles (3.04x 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 fully vectorizing your function, you can lower the cost of an iteration from 10.13 to 0.43 cycles (23.56x speedup).

Details

18% of SSE/AVX instructions are used in vector version (process two or more data elements in vector registers):

25% of SSE/AVX loads are used in vector version.
22% of SSE/AVX stores are used in vector version.
0% of SSE/AVX addition or subtraction 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.

omp_get_level@plt: 1 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:53
qmcplusplus::TimerManager
>::push_timer(qmcplusplus::TimerType
*): 1 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:73
std::chrono::_V2::system_clock::now(): 1 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:75

Vector unaligned load/store instructions

Detected 3 suboptimal vector unaligned load/store instructions.

Details

MOVUPS: 3 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:64
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:64
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:69

Workaround

Pass to your compiler a micro-architecture specialization option:
- use march=native
Use vector aligned instructions:
1. align your arrays on 64 bytes boundaries: replace { void *p = malloc (size); } with { void *p; posix_memalign (&p, 64, size); }.
2. 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

MOVZX: 2 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:65
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:70

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

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 91 bytes. The binary function is storing 67 bytes.

General properties

nb instructions	81
nb uops	81
loop length	284
used x86 registers	11
used mmx registers	0
used xmm registers	1
used ymm registers	0
used zmm registers	0
nb stack references	2

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	10.13 cycles
front end	10.13 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15
uops	6.17	6.17	6.17	6.17	6.17	6.17	7.25	7.25	7.25	7.25	0.00	0.00	0.00	0.00	1.00	1.00
cycles	6.17	6.17	6.17	6.17	6.17	6.17	7.25	7.25	7.25	7.25	0.00	0.00	0.00	0.00	1.00	1.00

Execution ports to units layout:

P0: ALU
P1: ALU
P2: ALU
P3: ALU, BRU
P4: ALU, BRU
P5: ALU, BRU
P6 (512 bits): store address, load
P7 (512 bits): store address, load
P8 (512 bits): store address, load
P9 (512 bits): store address, load
P10 (512 bits): VPU, DIV/SQRT
P11 (512 bits): VPU, DIV/SQRT
P12 (512 bits): VPU
P13 (512 bits): VPU
P14 (512 bits): FP store data
P15 (512 bits): FP store data

Cycles executing div or sqrt instructions

Cycles summary

Front-end	10.13
Dispatch	7.25
Overall L1	10.13

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	100%
load	100%
store	100%
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	100%

INT+FP

all	18%
load	25%
store	22%
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	10%

Vector efficiency ratios

INT

all	7%
load	6%
store	7%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	6%
fma	NA (no fma vectorizable/vectorized instructions)
other	8%

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

INT+FP

all	10%
load	11%
store	11%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	6%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	10%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 10.13 cycles. At this rate:

3% of peak load performance is reached (8.99 out of 256.00 bytes loaded per cycle (GB/s @ 1GHz))
5% of peak store performance is reached (6.62 out of 128.00 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 10.13 to 7.25 cycles (1.40x speedup).

ASM code

In the binary file, the address of the function is: 4a9a00

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P14	P15	Latency	Recip. throughput	Vectorization
CMPB $0,0x38(%RDI)	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	1	0.25	scal (1.6%)
JE 4a9bb8 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x1b8>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
PUSH %RBP	1	0	0	0	0	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	0	0	0	0.17	N/A
PUSH %R15	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R14	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R13	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R12	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RBX	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB $0x18,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
MOV %RDI,%R15	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
CALL 405970 <omp_get_level@plt>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0	0	0.50	N/A
TEST %EAX,%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	scal (6.3%)
JLE 4a9b2c <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x12c>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
MOV %R15,-0x38(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	scal (12.5%)
MOV $0x1,%R14B	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
MOV %EAX,-0x2c(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	scal (6.3%)
CMP $0x8,%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	scal (6.3%)
JB 4a9add <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0xdd>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
MOV -0x2c(%RBP),%EAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
MOV %EAX,%ECX	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
SHR $0x3,%ECX	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.33	N/A
MOV $0x1,%R14B	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
MOV %EAX,%R15D	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
NOPL (%RAX)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
MOV -0x2c(%RBP),%R12D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	scal (6.3%)
MOV %R12D,%EAX	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
AND $-0x8,%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
SUB %EAX,%R12D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	scal (6.3%)
JBE 4a9b22 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x122>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
XOR %EBX,%EBX	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.25	scal (6.3%)
MOV -0x38(%RBP),%R15	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
NOPW %CS:(%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
TEST $0x1,%R14B	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
JNE 4a9b2c <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x12c>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
JMP 4a9baa <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x1aa>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	1	N/A
MOV -0x38(%RBP),%R15	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
TEST $0x1,%R14B	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
JE 4a9baa <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x1aa>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
MOV 0x48(%R15),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
TEST %RAX,%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	scal (12.5%)
JE 4a9ba2 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x1a2>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
MOV 0x48(%RAX),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
CMP 0x40(%RAX),%RCX	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	1	0.25	scal (12.5%)
JE 4a9b75 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x175>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
MOV -0x8(%RCX),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
TEST %RAX,%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	scal (12.5%)
JE 4a9b75 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x175>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
MOVUPS 0x50(%RAX),%XMM0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.50	vect (25.0%)
MOVUPS %XMM0,0x50(%R15)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.50	0.50	4	0.50	vect (25.0%)
MOVSXD 0x60(%RAX),%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	1	0.25	N/A
MOV %EAX,0x60(%R15)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	scal (6.3%)
MOVZX 0x40(%R15),%ECX	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	1	0.25	N/A
MOV %CL,0x50(%R15,%RAX,1)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	N/A
MOV 0x60(%R15),%EAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
CMP $0xf,%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	scal (6.3%)
JL 4a9b90 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x190>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
MOVB $0x1,0x102864(%RIP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	scal (1.6%)
JMP 4a9b96 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x196>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	1	N/A
XORPS %XMM0,%XMM0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
MOVUPS %XMM0,0x50(%R15)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.50	0.50	4	0.50	vect (25.0%)
MOVZX 0x40(%R15),%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	1	0.25	N/A
MOV %AL,0x50(%R15)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	N/A
MOVL $0x1,0x60(%R15)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	scal (6.3%)
JMP 4a9b96 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x196>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	1	N/A
INC %EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
MOV %EAX,0x60(%R15)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	scal (6.3%)
MOV 0x48(%R15),%RDI	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
MOV %R15,%RSI	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	scal (12.5%)
CALL 4af8d0 <_ZN11qmcplusplus12TimerManagerINS_9TimerTypeINSt6chrono3_V212system_clockEEEE10push_timerEPS5_>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0	0	0.50	N/A
CALL 405a60 <_ZNSt6chrono3_V212system_clock3nowEv@plt>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0	0	0.50	N/A
MOV %RAX,(%R15)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	scal (12.5%)
ADD $0x18,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
POP %RBX	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R12	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R13	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R14	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R15	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %RBP	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
RET	1	0	0	0	0.33	0.33	0.33	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	0.09	N/A

The code analyzed by CQA in that panel excludes loops and represents 0.01% of application time for run icx_3

Source file and lines	NewTimer.cpp:39-81
Module	exec

The function is defined in:

/home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/TimerManager.h:92-93
/home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:39,53-75,81
/home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.h:90-97
/usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_vector.h:919

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 10.13 to 3.33 cycles (3.04x speedup).

Workaround

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

Vectorization

Details

18% of SSE/AVX instructions are used in vector version (process two or more data elements in vector registers):

25% of SSE/AVX loads are used in vector version.
22% of SSE/AVX stores are used in vector version.
0% of SSE/AVX addition or subtraction 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

omp_get_level@plt: 1 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:53
qmcplusplus::TimerManager
>::push_timer(qmcplusplus::TimerType
*): 1 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:73
std::chrono::_V2::system_clock::now(): 1 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:75

Vector unaligned load/store instructions

Detected 3 suboptimal vector unaligned load/store instructions.

Details

MOVUPS: 3 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:64
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:64
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:69

Workaround

Pass to your compiler a micro-architecture specialization option:
- use march=native
Use vector aligned instructions:
1. align your arrays on 64 bytes boundaries: replace { void *p = malloc (size); } with { void *p; posix_memalign (&p, 64, size); }.
2. 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

MOVZX: 2 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:65
- /home/eoseret/qaas_runs_ZEN5/173-938-0042/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:70

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

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 91 bytes. The binary function is storing 67 bytes.

General properties

nb instructions	81
nb uops	81
loop length	284
used x86 registers	11
used mmx registers	0
used xmm registers	1
used ymm registers	0
used zmm registers	0
nb stack references	2

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	10.13 cycles
front end	10.13 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15
uops	6.17	6.17	6.17	6.17	6.17	6.17	7.25	7.25	7.25	7.25	0.00	0.00	0.00	0.00	1.00	1.00
cycles	6.17	6.17	6.17	6.17	6.17	6.17	7.25	7.25	7.25	7.25	0.00	0.00	0.00	0.00	1.00	1.00

Execution ports to units layout:

P0: ALU
P1: ALU
P2: ALU
P3: ALU, BRU
P4: ALU, BRU
P5: ALU, BRU
P6 (512 bits): store address, load
P7 (512 bits): store address, load
P8 (512 bits): store address, load
P9 (512 bits): store address, load
P10 (512 bits): VPU, DIV/SQRT
P11 (512 bits): VPU, DIV/SQRT
P12 (512 bits): VPU
P13 (512 bits): VPU
P14 (512 bits): FP store data
P15 (512 bits): FP store data

Cycles executing div or sqrt instructions

Cycles summary

Front-end	10.13
Dispatch	7.25
Overall L1	10.13

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	100%
load	100%
store	100%
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	100%

INT+FP

all	18%
load	25%
store	22%
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	10%

Vector efficiency ratios

INT

all	7%
load	6%
store	7%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	6%
fma	NA (no fma vectorizable/vectorized instructions)
other	8%

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

INT+FP

all	10%
load	11%
store	11%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	6%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	10%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 10.13 cycles. At this rate:

3% of peak load performance is reached (8.99 out of 256.00 bytes loaded per cycle (GB/s @ 1GHz))
5% of peak store performance is reached (6.62 out of 128.00 bytes stored per cycle (GB/s @ 1GHz))

Front-end bottlenecks

ASM code

In the binary file, the address of the function is: 4a9a00

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P14	P15	Latency	Recip. throughput	Vectorization
CMPB $0,0x38(%RDI)	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	1	0.25	scal (1.6%)
JE 4a9bb8 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x1b8>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
PUSH %RBP	1	0	0	0	0	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	0	0	0	0.17	N/A
PUSH %R15	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R14	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R13	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R12	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RBX	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB $0x18,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
MOV %RDI,%R15	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
CALL 405970 <omp_get_level@plt>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0	0	0.50	N/A
TEST %EAX,%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	scal (6.3%)
JLE 4a9b2c <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x12c>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
MOV %R15,-0x38(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	scal (12.5%)
MOV $0x1,%R14B	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
MOV %EAX,-0x2c(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	scal (6.3%)
CMP $0x8,%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	scal (6.3%)
JB 4a9add <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0xdd>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
MOV -0x2c(%RBP),%EAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
MOV %EAX,%ECX	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
SHR $0x3,%ECX	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.33	N/A
MOV $0x1,%R14B	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
MOV %EAX,%R15D	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
NOPL (%RAX)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
MOV -0x2c(%RBP),%R12D	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	scal (6.3%)
MOV %R12D,%EAX	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
AND $-0x8,%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
SUB %EAX,%R12D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	scal (6.3%)
JBE 4a9b22 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x122>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
XOR %EBX,%EBX	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.25	scal (6.3%)
MOV -0x38(%RBP),%R15	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
NOPW %CS:(%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
TEST $0x1,%R14B	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
JNE 4a9b2c <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x12c>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
JMP 4a9baa <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x1aa>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	1	N/A
MOV -0x38(%RBP),%R15	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
TEST $0x1,%R14B	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
JE 4a9baa <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x1aa>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
MOV 0x48(%R15),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
TEST %RAX,%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	scal (12.5%)
JE 4a9ba2 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x1a2>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
MOV 0x48(%RAX),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
CMP 0x40(%RAX),%RCX	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	1	0.25	scal (12.5%)
JE 4a9b75 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x175>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
MOV -0x8(%RCX),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
TEST %RAX,%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	scal (12.5%)
JE 4a9b75 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x175>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
MOVUPS 0x50(%RAX),%XMM0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.50	vect (25.0%)
MOVUPS %XMM0,0x50(%R15)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.50	0.50	4	0.50	vect (25.0%)
MOVSXD 0x60(%RAX),%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	1	0.25	N/A
MOV %EAX,0x60(%R15)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	scal (6.3%)
MOVZX 0x40(%R15),%ECX	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	1	0.25	N/A
MOV %CL,0x50(%R15,%RAX,1)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	N/A
MOV 0x60(%R15),%EAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
CMP $0xf,%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	scal (6.3%)
JL 4a9b90 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x190>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	1	0.50-1	N/A
MOVB $0x1,0x102864(%RIP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	scal (1.6%)
JMP 4a9b96 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x196>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	1	N/A
XORPS %XMM0,%XMM0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	vect (25.0%)
MOVUPS %XMM0,0x50(%R15)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0.50	0.50	4	0.50	vect (25.0%)
MOVZX 0x40(%R15),%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0.25	0.25	0.25	0.25	0	0	1	0.25	N/A
MOV %AL,0x50(%R15)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	N/A
MOVL $0x1,0x60(%R15)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	scal (6.3%)
JMP 4a9b96 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv+0x196>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	0	0	1	N/A
INC %EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
MOV %EAX,0x60(%R15)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	scal (6.3%)
MOV 0x48(%R15),%RDI	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	3	0.25	N/A
MOV %R15,%RSI	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	scal (12.5%)
CALL 4af8d0 <_ZN11qmcplusplus12TimerManagerINS_9TimerTypeINSt6chrono3_V212system_clockEEEE10push_timerEPS5_>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0	0	0.50	N/A
CALL 405a60 <_ZNSt6chrono3_V212system_clock3nowEv@plt>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0	0	0.50	N/A
MOV %RAX,(%R15)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	4	0.50	scal (12.5%)
ADD $0x18,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	0	0	1	0.17	N/A
POP %RBX	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R12	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R13	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R14	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R15	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %RBP	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
RET	1	0	0	0	0.33	0.33	0.33	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	0.09	N/A

Name	Coverage (%)	Time (s)
▼qmcplusplus::TimerType::start()–	0.01	0.02
○Loop 1702 - NewTimer.cpp:53-54 - exec	0.00	0.01
○Loop 1703 - NewTimer.cpp:39-54 - exec	0.00	0.00

Report Configuration

Code clean check

Workaround

Vectorization

Details

Workaround

Execution units bottlenecks

CALL 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

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

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

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code