OV - exec - Function qmcplusplus::SoaDistanceTableABOMPTarget<double, 3u, 40>::update(int)

	Function: qmcplusplus::SoaDistanceTableABOMPTarget<double, 3u, 40>::update(int)	Module: exec	Source: SoaDistanceTableABOMPTarget.h:372-377 [...]	Coverage (incl. loops): 0.02% \| (excl. loops): 0.02%

/home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/VectorSoAContainer.h: 271 - 271

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

271:   inline T* restrict data(size_t i) { return myData + i * nGhosts; }

/home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.h: 242 - 249

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

242:   ScopeGuard(TIMER& t) : timer(t) { timer.start(); }

[...]

249:   ~ScopeGuard() { timer.stop(); }

/usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_vector.h: 1046 - 1046

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

1046: 	return *(this->_M_impl._M_start + __n);

/usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algobase.h: 429 - 431

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

429: 	  const ptrdiff_t _Num = __last - __first;

430: 	  if (_Num)

431: 	    __builtin_memmove(__result, __first, sizeof(_Tp) * _Num);

/usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algo.h: 751 - 751

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

751:       if (__n2 <= 0)

/home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsVector.h: 248 - 248

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

248:   inline pointer data() { return X; }

/home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Particle/SoaDistanceTableABOMPTarget.h: 372 - 377

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

372:   {

373:     ScopedTimer local_timer(update_timer_);

374:     std::copy_n(temp_r_.data(), num_sources_, distances_[iat].data());

375:     for (int idim = 0; idim < D; ++idim)

376:       std::copy_n(temp_dr_.data(idim), num_sources_, displacements_[iat].data(idim));

377:   }

0x4ce400 PUSH	%RBP

0x4ce401 MOV	%RSP,%RBP

0x4ce404 PUSH	%R15

0x4ce406 PUSH	%R14

0x4ce408 PUSH	%RBX

0x4ce409 PUSH	%RAX

0x4ce40a MOV	%ESI,%R15D

0x4ce40d MOV	%RDI,%R14

0x4ce410 MOV	0x1c8(%RDI),%RBX

0x4ce417 MOV	%RBX,%RDI

0x4ce41a CALL	48dfa0 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv>

0x4ce41f MOV	0x10(%R14),%RDX

0x4ce423 TEST	%RDX,%RDX

0x4ce426 JE	4ce4e5

0x4ce42c MOVSXD	%R15D,%RAX

0x4ce42f LEA	(%RAX,%RAX,4),%R15

0x4ce433 MOV	0x48(%R14),%RAX

0x4ce437 MOV	0x90(%R14),%RSI

0x4ce43e MOV	0x18(%RAX,%R15,8),%RDI

0x4ce443 SAL	$0x3,%RDX

0x4ce447 CALL	4040b0 <memmove@plt>

0x4ce44c CMPQ	$0,0x10(%R14)

0x4ce451 JE	4ce4e5

0x4ce457 MOV	0x10(%R14),%RDX

0x4ce45b TEST	%RDX,%RDX

0x4ce45e JE	4ce479

0x4ce460 MOV	0x60(%R14),%RAX

0x4ce464 MOV	0xb8(%R14),%RSI

0x4ce46b MOV	0x18(%RAX,%R15,8),%RDI

0x4ce470 SAL	$0x3,%RDX

0x4ce474 CALL	4040b0 <memmove@plt>

0x4ce479 MOV	0x10(%R14),%RDX

0x4ce47d TEST	%RDX,%RDX

0x4ce480 JE	4ce4af

0x4ce482 MOV	0x60(%R14),%RAX

0x4ce486 MOV	0xa8(%R14),%RSI

0x4ce48d MOV	0x8(%RAX,%R15,8),%RDI

0x4ce492 SAL	$0x3,%RDI

0x4ce496 ADD	0x18(%RAX,%R15,8),%RDI

0x4ce49b SAL	$0x3,%RSI

0x4ce49f ADD	0xb8(%R14),%RSI

0x4ce4a6 SAL	$0x3,%RDX

0x4ce4aa CALL	4040b0 <memmove@plt>

0x4ce4af MOV	0x10(%R14),%RDX

0x4ce4b3 TEST	%RDX,%RDX

0x4ce4b6 JE	4ce4e5

0x4ce4b8 MOV	0x60(%R14),%RAX

0x4ce4bc MOV	0xa8(%R14),%RSI

0x4ce4c3 MOV	0x8(%RAX,%R15,8),%RDI

0x4ce4c8 SAL	$0x4,%RDI

0x4ce4cc ADD	0x18(%RAX,%R15,8),%RDI

0x4ce4d1 SAL	$0x4,%RSI

0x4ce4d5 ADD	0xb8(%R14),%RSI

0x4ce4dc SAL	$0x3,%RDX

0x4ce4e0 CALL	4040b0 <memmove@plt>

0x4ce4e5 MOV	%RBX,%RDI

0x4ce4e8 CALL	48e150 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE4stopEv>

0x4ce4ed ADD	$0x8,%RSP

0x4ce4f1 POP	%RBX

0x4ce4f2 POP	%R14

0x4ce4f4 POP	%R15

0x4ce4f6 POP	%RBP

0x4ce4f7 RET

0x4ce4f8 MOV	%RAX,%RDI

0x4ce4fb CALL	40e360 <__clang_call_terminate>

Coverage (%)	Name	Source Location	Module
►78.79+	qmcplusplus::ParticleSet::acce[...]	ParticleSet.cpp:389	exec
○	main.extracted.107	miniqmc.cpp:452	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so
►21.21+	main.extracted.107	miniqmc.cpp:452	exec
○	__kmp_invoke_microtask		libiomp5.so
○	__kmp_invoke_task_func		libiomp5.so

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 0.02% of application time for run orig_0

Source file and lines	SoaDistanceTableABOMPTarget.h:372-377
Module	exec

The function is defined in:

/home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/VectorSoAContainer.h:271
/home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.h:242,249
/usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_vector.h:1046
/usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algobase.h:429-431
/usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algo.h:751
/home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsVector.h:248
/home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Particle/SoaDistanceTableABOMPTarget.h:372-377

Warnings:

Ignoring paths for analysis
Failed to get the number of paths

gain
potential
hint
expert

Vectorization

Your function is not vectorized. 8 data elements could be processed at once in vector registers. By vectorizing your function, you can lower the cost of an iteration from 12.33 to 1.54 cycles (8.00x 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.

__clang_call_terminate: 1 occurrences►
- /home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.h:249
memmove@plt: 4 occurrences►
- /usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algobase.h:431
- /usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algobase.h:431
- /usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algobase.h:431
- /usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algobase.h:431
qmcplusplus::TimerType
::start(): 1 occurrences►
- /home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.h:242
qmcplusplus::TimerType
::stop(): 1 occurrences►
- /home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.h:249

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 176 bytes.

General properties

nb instructions	65
nb uops	72
loop length	256
used x86 registers	9
used mmx registers	0
used xmm registers	0
used ymm registers	0
used zmm registers	0
nb stack references	0

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	12.33 cycles
front end	12.33 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11
uops	7.00	3.33	9.00	9.00	6.00	3.33	7.00	6.00	6.00	6.00	3.33	9.00
cycles	7.00	3.33	9.00	9.00	6.00	3.33	7.00	6.00	6.00	6.00	3.33	9.00

Execution ports to units layout:

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

Cycles executing div or sqrt instructions

Cycles summary

Front-end	12.33
Dispatch	9.00
Overall L1	12.33

Vectorization ratios

all	0%
load	0%
store	NA (no store vectorizable/vectorized instructions)
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	12%
load	12%
store	NA (no store vectorizable/vectorized instructions)
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 12.33 cycles. At this rate:

7% of peak load performance is reached (14.27 out of 192.00 bytes loaded 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 12.33 to 9.00 cycles (1.37x speedup).

ASM code

In the binary file, the address of the function is: 4ce400

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	Latency	Recip. throughput	Vectorization
PUSH %RBP	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	5-12	0.50	N/A
MOV %RSP,%RBP	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.17	scal (12.5%)
PUSH %R15	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	5-12	0.50	N/A
PUSH %R14	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	5-12	0.50	N/A
PUSH %RBX	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	5-12	0.50	N/A
PUSH %RAX	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	5-12	0.50	N/A
MOV %ESI,%R15D	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
MOV %RDI,%R14	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
MOV 0x1c8(%RDI),%RBX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
MOV %RBX,%RDI	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
CALL 48dfa0 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv>	2	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0	1	N/A
MOV 0x10(%R14),%RDX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
TEST %RDX,%RDX	1	0.20	0.20	0	0	0	0.20	0.20	0	0	0	0.20	0	2	0.20	scal (12.5%)
JE 4ce4e5 <_ZN11qmcplusplus27SoaDistanceTableABOMPTargetIdLj3ELi40EE6updateEi+0xe5>	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0	0.50	N/A
MOVSXD %R15D,%RAX	1	0	0.33	0	0	0	0.33	0	0	0	0	0.33	0	1	0.33	N/A
LEA (%RAX,%RAX,4),%R15	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV 0x48(%R14),%RAX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
MOV 0x90(%R14),%RSI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
MOV 0x18(%RAX,%R15,8),%RDI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
SAL $0x3,%RDX	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	scal (12.5%)
CALL 4040b0 <memmove@plt>	2	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0	1	N/A
CMPQ $0,0x10(%R14)	1	0.20	0.20	0.33	0.33	0	0.20	0.20	0	0	0	0.20	0.33	1	0.33	scal (12.5%)
JE 4ce4e5 <_ZN11qmcplusplus27SoaDistanceTableABOMPTargetIdLj3ELi40EE6updateEi+0xe5>	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0	0.50	N/A
MOV 0x10(%R14),%RDX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
TEST %RDX,%RDX	1	0.20	0.20	0	0	0	0.20	0.20	0	0	0	0.20	0	2	0.20	scal (12.5%)
JE 4ce479 <_ZN11qmcplusplus27SoaDistanceTableABOMPTargetIdLj3ELi40EE6updateEi+0x79>	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0	0.50	N/A
MOV 0x60(%R14),%RAX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
MOV 0xb8(%R14),%RSI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
MOV 0x18(%RAX,%R15,8),%RDI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
SAL $0x3,%RDX	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	scal (12.5%)
CALL 4040b0 <memmove@plt>	2	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0	1	N/A
MOV 0x10(%R14),%RDX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
TEST %RDX,%RDX	1	0.20	0.20	0	0	0	0.20	0.20	0	0	0	0.20	0	2	0.20	scal (12.5%)
JE 4ce4af <_ZN11qmcplusplus27SoaDistanceTableABOMPTargetIdLj3ELi40EE6updateEi+0xaf>	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0	0.50	N/A
MOV 0x60(%R14),%RAX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
MOV 0xa8(%R14),%RSI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
MOV 0x8(%RAX,%R15,8),%RDI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
SAL $0x3,%RDI	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	N/A
ADD 0x18(%RAX,%R15,8),%RDI	1	0.20	0.20	0.33	0.33	0	0.20	0.20	0	0	0	0.20	0.33	1	0.33	N/A
SAL $0x3,%RSI	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	scal (12.5%)
ADD 0xb8(%R14),%RSI	1	0.20	0.20	0.33	0.33	0	0.20	0.20	0	0	0	0.20	0.33	1	0.33	scal (12.5%)
SAL $0x3,%RDX	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	scal (12.5%)
CALL 4040b0 <memmove@plt>	2	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0	1	N/A
MOV 0x10(%R14),%RDX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
TEST %RDX,%RDX	1	0.20	0.20	0	0	0	0.20	0.20	0	0	0	0.20	0	2	0.20	scal (12.5%)
JE 4ce4e5 <_ZN11qmcplusplus27SoaDistanceTableABOMPTargetIdLj3ELi40EE6updateEi+0xe5>	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0	0.50	N/A
MOV 0x60(%R14),%RAX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
MOV 0xa8(%R14),%RSI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
MOV 0x8(%RAX,%R15,8),%RDI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
SAL $0x4,%RDI	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	N/A
ADD 0x18(%RAX,%R15,8),%RDI	1	0.20	0.20	0.33	0.33	0	0.20	0.20	0	0	0	0.20	0.33	1	0.33	N/A
SAL $0x4,%RSI	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	scal (12.5%)
ADD 0xb8(%R14),%RSI	1	0.20	0.20	0.33	0.33	0	0.20	0.20	0	0	0	0.20	0.33	1	0.33	scal (12.5%)
SAL $0x3,%RDX	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	scal (12.5%)
CALL 4040b0 <memmove@plt>	2	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0	1	N/A
MOV %RBX,%RDI	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
CALL 48e150 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE4stopEv>	2	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0	1	N/A
ADD $0x8,%RSP	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
POP %RBX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1-6	0.33	N/A
POP %R14	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1-6	0.33	N/A
POP %R15	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1-6	0.33	N/A
POP %RBP	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1-6	0.33	N/A
RET	1	0.50	0	0.33	0.33	0	0	0.50	0	0	0	0	0.33	0	2.13	N/A
MOV %RAX,%RDI	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
CALL 40e360 <__clang_call_terminate>	2	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0	1	N/A

The code analyzed by CQA in that panel excludes loops and represents 0.02% of application time for run orig_0

Source file and lines	SoaDistanceTableABOMPTarget.h:372-377
Module	exec

The function is defined in:

/home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/VectorSoAContainer.h:271
/home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.h:242,249
/usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_vector.h:1046
/usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algobase.h:429-431
/usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algo.h:751
/home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsVector.h:248
/home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Particle/SoaDistanceTableABOMPTarget.h:372-377

Warnings:

Ignoring paths for analysis
Failed to get the number of paths

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

__clang_call_terminate: 1 occurrences►
- /home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.h:249
memmove@plt: 4 occurrences►
- /usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algobase.h:431
- /usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algobase.h:431
- /usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algobase.h:431
- /usr/lib/gcc/x86_64-redhat-linux/11/../../../../include/c++/11/bits/stl_algobase.h:431
qmcplusplus::TimerType
::start(): 1 occurrences►
- /home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.h:242
qmcplusplus::TimerType
::stop(): 1 occurrences►
- /home/eoseret/qaas_runs_GNR/173-773-5521/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.h:249

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 176 bytes.

General properties

nb instructions	65
nb uops	72
loop length	256
used x86 registers	9
used mmx registers	0
used xmm registers	0
used ymm registers	0
used zmm registers	0
nb stack references	0

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	12.33 cycles
front end	12.33 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11
uops	7.00	3.33	9.00	9.00	6.00	3.33	7.00	6.00	6.00	6.00	3.33	9.00
cycles	7.00	3.33	9.00	9.00	6.00	3.33	7.00	6.00	6.00	6.00	3.33	9.00

Execution ports to units layout:

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

Cycles executing div or sqrt instructions

Cycles summary

Front-end	12.33
Dispatch	9.00
Overall L1	12.33

Vectorization ratios

all	0%
load	0%
store	NA (no store vectorizable/vectorized instructions)
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	12%
load	12%
store	NA (no store vectorizable/vectorized instructions)
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 12.33 cycles. At this rate:

7% of peak load performance is reached (14.27 out of 192.00 bytes loaded per cycle (GB/s @ 1GHz))

Front-end bottlenecks

ASM code

In the binary file, the address of the function is: 4ce400

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	Latency	Recip. throughput	Vectorization
PUSH %RBP	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	5-12	0.50	N/A
MOV %RSP,%RBP	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.17	scal (12.5%)
PUSH %R15	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	5-12	0.50	N/A
PUSH %R14	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	5-12	0.50	N/A
PUSH %RBX	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	5-12	0.50	N/A
PUSH %RAX	1	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	5-12	0.50	N/A
MOV %ESI,%R15D	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
MOV %RDI,%R14	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
MOV 0x1c8(%RDI),%RBX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
MOV %RBX,%RDI	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
CALL 48dfa0 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE5startEv>	2	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0	1	N/A
MOV 0x10(%R14),%RDX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
TEST %RDX,%RDX	1	0.20	0.20	0	0	0	0.20	0.20	0	0	0	0.20	0	2	0.20	scal (12.5%)
JE 4ce4e5 <_ZN11qmcplusplus27SoaDistanceTableABOMPTargetIdLj3ELi40EE6updateEi+0xe5>	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0	0.50	N/A
MOVSXD %R15D,%RAX	1	0	0.33	0	0	0	0.33	0	0	0	0	0.33	0	1	0.33	N/A
LEA (%RAX,%RAX,4),%R15	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV 0x48(%R14),%RAX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
MOV 0x90(%R14),%RSI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
MOV 0x18(%RAX,%R15,8),%RDI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
SAL $0x3,%RDX	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	scal (12.5%)
CALL 4040b0 <memmove@plt>	2	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0	1	N/A
CMPQ $0,0x10(%R14)	1	0.20	0.20	0.33	0.33	0	0.20	0.20	0	0	0	0.20	0.33	1	0.33	scal (12.5%)
JE 4ce4e5 <_ZN11qmcplusplus27SoaDistanceTableABOMPTargetIdLj3ELi40EE6updateEi+0xe5>	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0	0.50	N/A
MOV 0x10(%R14),%RDX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
TEST %RDX,%RDX	1	0.20	0.20	0	0	0	0.20	0.20	0	0	0	0.20	0	2	0.20	scal (12.5%)
JE 4ce479 <_ZN11qmcplusplus27SoaDistanceTableABOMPTargetIdLj3ELi40EE6updateEi+0x79>	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0	0.50	N/A
MOV 0x60(%R14),%RAX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
MOV 0xb8(%R14),%RSI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
MOV 0x18(%RAX,%R15,8),%RDI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
SAL $0x3,%RDX	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	scal (12.5%)
CALL 4040b0 <memmove@plt>	2	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0	1	N/A
MOV 0x10(%R14),%RDX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
TEST %RDX,%RDX	1	0.20	0.20	0	0	0	0.20	0.20	0	0	0	0.20	0	2	0.20	scal (12.5%)
JE 4ce4af <_ZN11qmcplusplus27SoaDistanceTableABOMPTargetIdLj3ELi40EE6updateEi+0xaf>	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0	0.50	N/A
MOV 0x60(%R14),%RAX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
MOV 0xa8(%R14),%RSI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
MOV 0x8(%RAX,%R15,8),%RDI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
SAL $0x3,%RDI	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	N/A
ADD 0x18(%RAX,%R15,8),%RDI	1	0.20	0.20	0.33	0.33	0	0.20	0.20	0	0	0	0.20	0.33	1	0.33	N/A
SAL $0x3,%RSI	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	scal (12.5%)
ADD 0xb8(%R14),%RSI	1	0.20	0.20	0.33	0.33	0	0.20	0.20	0	0	0	0.20	0.33	1	0.33	scal (12.5%)
SAL $0x3,%RDX	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	scal (12.5%)
CALL 4040b0 <memmove@plt>	2	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0	1	N/A
MOV 0x10(%R14),%RDX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
TEST %RDX,%RDX	1	0.20	0.20	0	0	0	0.20	0.20	0	0	0	0.20	0	2	0.20	scal (12.5%)
JE 4ce4e5 <_ZN11qmcplusplus27SoaDistanceTableABOMPTargetIdLj3ELi40EE6updateEi+0xe5>	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0	0.50	N/A
MOV 0x60(%R14),%RAX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
MOV 0xa8(%R14),%RSI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	scal (12.5%)
MOV 0x8(%RAX,%R15,8),%RDI	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1	0.33	N/A
SAL $0x4,%RDI	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	N/A
ADD 0x18(%RAX,%R15,8),%RDI	1	0.20	0.20	0.33	0.33	0	0.20	0.20	0	0	0	0.20	0.33	1	0.33	N/A
SAL $0x4,%RSI	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	scal (12.5%)
ADD 0xb8(%R14),%RSI	1	0.20	0.20	0.33	0.33	0	0.20	0.20	0	0	0	0.20	0.33	1	0.33	scal (12.5%)
SAL $0x3,%RDX	1	0.50	0	0	0	0	0	0.50	0	0	0	0	0	0-2	0.50	scal (12.5%)
CALL 4040b0 <memmove@plt>	2	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0	1	N/A
MOV %RBX,%RDI	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
CALL 48e150 <_ZN11qmcplusplus9TimerTypeINSt6chrono3_V212system_clockEE4stopEv>	2	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0	1	N/A
ADD $0x8,%RSP	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
POP %RBX	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1-6	0.33	N/A
POP %R14	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1-6	0.33	N/A
POP %R15	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1-6	0.33	N/A
POP %RBP	1	0	0	0.33	0.33	0	0	0	0	0	0	0	0.33	1-6	0.33	N/A
RET	1	0.50	0	0.33	0.33	0	0	0.50	0	0	0	0	0.33	0	2.13	N/A
MOV %RAX,%RDI	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0.17	N/A
CALL 40e360 <__clang_call_terminate>	2	0	0	0	0	0.50	0	0	0.50	0.50	0.50	0	0	0	1	N/A

Name	Coverage (%)	Time (s)
○qmcplusplus::SoaDistanceTableABOMPTarget::update(int)	0.02	0.01

Report Configuration

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

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