Loop Id: 17420 | Module: libgromacs_mpi.so.9.0.0 | Source: pme_redistribute.cpp:436-449 [...] | Coverage: NA% |
---|
Loop Id: 17420 | Module: libgromacs_mpi.so.9.0.0 | Source: pme_redistribute.cpp:436-449 [...] | Coverage: NA% |
---|
0xc0a300 MOV -0x50(%RBP),%R8 [5] |
0xc0a304 MOV 0x768(%R8),%R8 [7] |
0xc0a30b MOV 0x30(%R10),%R9 [1] |
0xc0a30f MOVSXD (%R9,%RDI,4),%R10 [2] |
0xc0a313 LEA (%R10,%R10,2),%R11 |
0xc0a317 VMOVSS (%RBX),%XMM0 [8] |
0xc0a31b VADDSS 0x8(%R8,%R11,4),%XMM0,%XMM0 [6] |
0xc0a322 VMOVSD -0x8(%RBX),%XMM1 [8] |
0xc0a327 VMOVSD (%R8,%R11,4),%XMM2 [6] |
0xc0a32d VADDPS %XMM2,%XMM1,%XMM1 |
0xc0a331 VMOVLPS %XMM1,-0x8(%RBX) [8] |
0xc0a336 VMOVSS %XMM0,(%RBX) [8] |
0xc0a33a LEA 0x1(%R10),%R8D |
0xc0a33e MOV -0x30(%RBP),%R10 [5] |
0xc0a342 MOV %R8D,(%R9,%RDI,4) [2] |
0xc0a346 INC %RSI |
0xc0a349 ADD $0xc,%RBX |
0xc0a34d CMP %RSI,%RCX |
0xc0a350 JE c0a469 |
0xc0a356 MOVSXD (%RAX,%RSI,4),%RDI [4] |
0xc0a35a CMP 0x8(%R10),%EDI [1] |
0xc0a35e JNE c0a300 |
0xc0a360 MOVSXD %EDX,%RDX |
0xc0a363 MOV 0xa0(%R10),%RDI [1] |
0xc0a36a LEA (%RDX,%RDX,2),%R8 |
0xc0a36e VMOVSS (%RBX),%XMM0 [8] |
0xc0a372 VADDSS 0x8(%RDI,%R8,4),%XMM0,%XMM0 [3] |
0xc0a379 VMOVSD -0x8(%RBX),%XMM1 [8] |
0xc0a37e VMOVSD (%RDI,%R8,4),%XMM2 [3] |
0xc0a384 VADDPS %XMM2,%XMM1,%XMM1 |
0xc0a388 VMOVLPS %XMM1,-0x8(%RBX) [8] |
0xc0a38d VMOVSS %XMM0,(%RBX) [8] |
0xc0a391 INC %EDX |
0xc0a393 JMP c0a346 |
/home/eoseret/gromacs-2024.2/src/gromacs/ewald/pme_redistribute.cpp: 436 - 449 |
-------------------------------------------------------------------------------- |
436: for (gmx::Index i = 0; i < f.ssize(); i++) |
437: { |
438: const int slabIndex = atc->pd[i]; |
439: if (slabIndex == atc->slabIndex) |
440: { |
441: /* Add from the local force array */ |
442: rvec_inc(f[i], atc->f[local_pos]); |
443: local_pos++; |
444: } |
445: else |
446: { |
447: /* Add from the receive buffer */ |
448: rvec_inc(f[i], pme->bufv[atc->bufferIndices[slabIndex]]); |
449: atc->bufferIndices[slabIndex]++; |
/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); |
/home/eoseret/gromacs-2024.2/api/legacy/include/gromacs/utility/arrayref.h: 82 - 82 |
-------------------------------------------------------------------------------- |
82: it_ += i; |
/home/eoseret/gromacs-2024.2/src/external/boost/stl_interfaces/iterator_interface.hpp: 305 - 305 |
-------------------------------------------------------------------------------- |
305: D retval = derived(); |
/home/eoseret/gromacs-2024.2/api/legacy/include/gromacs/math/vec.h: 141 - 147 |
-------------------------------------------------------------------------------- |
141: x = a[XX] + b[XX]; |
142: y = a[YY] + b[YY]; |
143: z = a[ZZ] + b[ZZ]; |
144: |
145: a[XX] = x; |
146: a[YY] = y; |
147: a[ZZ] = z; |
Path / |
Metric | Value |
---|---|
CQA speedup if no scalar integer | 1.44 |
CQA speedup if FP arith vectorized | 1.43 |
CQA speedup if fully vectorized | 5.49 |
CQA speedup if no inter-iteration dependency | NA |
CQA speedup if next bottleneck killed | 1.33 |
Bottlenecks | |
Function | dd_pmeredist_f(gmx_pme_t*, PmeAtomComm*, gmx::ArrayRef |
Source | pme_redistribute.cpp:436-449,stl_vector.h:1046-1046,arrayref.h:82-82,iterator_interface.hpp:305-305,vec.h:141-147 |
Source loop unroll info | not unrolled or unrolled with no peel/tail loop |
Source loop unroll confidence level | max |
Unroll/vectorization loop type | NA |
Unroll factor | NA |
CQA cycles | 3.83 |
CQA cycles if no scalar integer | 2.67 |
CQA cycles if FP arith vectorized | 2.69 |
CQA cycles if fully vectorized | 0.70 |
Front-end cycles | 3.25 |
P0 cycles | 2.13 |
P1 cycles | 2.13 |
P2 cycles | 2.00 |
P3 cycles | 2.00 |
P4 cycles | 1.25 |
P5 cycles | 3.83 |
P6 cycles | 3.83 |
P7 cycles | 3.83 |
P8 cycles | 0.00 |
P9 cycles | 0.00 |
P10 cycles | 1.00 |
P11 cycles | 1.00 |
P12 cycles | 1.00 |
P13 cycles | 1.00 |
DIV/SQRT cycles | 0.00 |
Inter-iter dependencies cycles | 2 |
FE+BE cycles (UFS) | NA |
Stall cycles (UFS) | NA |
Nb insns | 20.50 |
Nb uops | 19.50 |
Nb loads | 9.00 |
Nb stores | 2.50 |
Nb stack references | 1.00 |
FLOP/cycle | 1.30 |
Nb FLOP add-sub | 5.00 |
Nb FLOP mul | 0.00 |
Nb FLOP fma | 0.00 |
Nb FLOP div | 0.00 |
Nb FLOP rcp | 0.00 |
Nb FLOP sqrt | 0.00 |
Nb FLOP rsqrt | 0.00 |
Bytes/cycle | 17.67 |
Bytes prefetched | 0.00 |
Bytes loaded | 54.00 |
Bytes stored | 14.00 |
Stride 0 | 1.00 |
Stride 1 | 1.00 |
Stride n | 1.00 |
Stride unknown | 1.50 |
Stride indirect | 1.00 |
Vectorization ratio all | 22.22 |
Vectorization ratio load | 0.00 |
Vectorization ratio store | 41.67 |
Vectorization ratio mul | NA |
Vectorization ratio add_sub | 50.00 |
Vectorization ratio fma | NA |
Vectorization ratio div_sqrt | NA |
Vectorization ratio other | 0.00 |
Vector-efficiency ratio all | 11.11 |
Vector-efficiency ratio load | 9.06 |
Vector-efficiency ratio store | 8.85 |
Vector-efficiency ratio mul | NA |
Vector-efficiency ratio add_sub | 15.63 |
Vector-efficiency ratio fma | NA |
Vector-efficiency ratio div_sqrt | NA |
Vector-efficiency ratio other | 10.94 |
Metric | Value |
---|---|
CQA speedup if no scalar integer | 1.29 |
CQA speedup if FP arith vectorized | 1.20 |
CQA speedup if fully vectorized | 5.76 |
CQA speedup if no inter-iteration dependency | NA |
CQA speedup if next bottleneck killed | 1.33 |
Bottlenecks | micro-operation queue, P5, P6, P7, |
Function | dd_pmeredist_f(gmx_pme_t*, PmeAtomComm*, gmx::ArrayRef |
Source | pme_redistribute.cpp:436-449,stl_vector.h:1046-1046,arrayref.h:82-82,iterator_interface.hpp:305-305,vec.h:141-147 |
Source loop unroll info | not unrolled or unrolled with no peel/tail loop |
Source loop unroll confidence level | max |
Unroll/vectorization loop type | NA |
Unroll factor | NA |
CQA cycles | 3.00 |
CQA cycles if no scalar integer | 2.33 |
CQA cycles if FP arith vectorized | 2.50 |
CQA cycles if fully vectorized | 0.52 |
Front-end cycles | 3.00 |
P0 cycles | 2.25 |
P1 cycles | 2.25 |
P2 cycles | 2.00 |
P3 cycles | 2.00 |
P4 cycles | 1.50 |
P5 cycles | 3.00 |
P6 cycles | 3.00 |
P7 cycles | 3.00 |
P8 cycles | 0.00 |
P9 cycles | 0.00 |
P10 cycles | 1.00 |
P11 cycles | 1.00 |
P12 cycles | 1.00 |
P13 cycles | 1.00 |
DIV/SQRT cycles | 0.00 |
Inter-iter dependencies cycles | 2 |
FE+BE cycles (UFS) | NA |
Stall cycles (UFS) | NA |
Nb insns | 19.00 |
Nb uops | 18.00 |
Nb loads | 7.00 |
Nb stores | 2.00 |
Nb stack references | 0.00 |
FLOP/cycle | 1.67 |
Nb FLOP add-sub | 5.00 |
Nb FLOP mul | 0.00 |
Nb FLOP fma | 0.00 |
Nb FLOP div | 0.00 |
Nb FLOP rcp | 0.00 |
Nb FLOP sqrt | 0.00 |
Nb FLOP rsqrt | 0.00 |
Bytes/cycle | 17.33 |
Bytes prefetched | 0.00 |
Bytes loaded | 40.00 |
Bytes stored | 12.00 |
Stride 0 | 1.00 |
Stride 1 | 1.00 |
Stride n | 1.00 |
Stride unknown | 1.00 |
Stride indirect | 0.00 |
Vectorization ratio all | 22.22 |
Vectorization ratio load | 0.00 |
Vectorization ratio store | 50.00 |
Vectorization ratio mul | NA |
Vectorization ratio add_sub | 50.00 |
Vectorization ratio fma | NA |
Vectorization ratio div_sqrt | NA |
Vectorization ratio other | 0.00 |
Vector-efficiency ratio all | 11.11 |
Vector-efficiency ratio load | 8.75 |
Vector-efficiency ratio store | 9.38 |
Vector-efficiency ratio mul | NA |
Vector-efficiency ratio add_sub | 15.63 |
Vector-efficiency ratio fma | NA |
Vector-efficiency ratio div_sqrt | NA |
Vector-efficiency ratio other | 9.38 |
Metric | Value |
---|---|
CQA speedup if no scalar integer | 1.56 |
CQA speedup if FP arith vectorized | 1.62 |
CQA speedup if fully vectorized | 5.33 |
CQA speedup if no inter-iteration dependency | NA |
CQA speedup if next bottleneck killed | 1.33 |
Bottlenecks | P5, P6, P7, |
Function | dd_pmeredist_f(gmx_pme_t*, PmeAtomComm*, gmx::ArrayRef |
Source | pme_redistribute.cpp:436-449,stl_vector.h:1046-1046,arrayref.h:82-82,iterator_interface.hpp:305-305,vec.h:141-147 |
Source loop unroll info | not unrolled or unrolled with no peel/tail loop |
Source loop unroll confidence level | max |
Unroll/vectorization loop type | NA |
Unroll factor | NA |
CQA cycles | 4.67 |
CQA cycles if no scalar integer | 3.00 |
CQA cycles if FP arith vectorized | 2.88 |
CQA cycles if fully vectorized | 0.88 |
Front-end cycles | 3.50 |
P0 cycles | 2.00 |
P1 cycles | 2.00 |
P2 cycles | 2.00 |
P3 cycles | 2.00 |
P4 cycles | 1.00 |
P5 cycles | 4.67 |
P6 cycles | 4.67 |
P7 cycles | 4.67 |
P8 cycles | 0.00 |
P9 cycles | 0.00 |
P10 cycles | 1.00 |
P11 cycles | 1.00 |
P12 cycles | 1.00 |
P13 cycles | 1.00 |
DIV/SQRT cycles | 0.00 |
Inter-iter dependencies cycles | 2 |
FE+BE cycles (UFS) | NA |
Stall cycles (UFS) | NA |
Nb insns | 22.00 |
Nb uops | 21.00 |
Nb loads | 11.00 |
Nb stores | 3.00 |
Nb stack references | 2.00 |
FLOP/cycle | 1.07 |
Nb FLOP add-sub | 5.00 |
Nb FLOP mul | 0.00 |
Nb FLOP fma | 0.00 |
Nb FLOP div | 0.00 |
Nb FLOP rcp | 0.00 |
Nb FLOP sqrt | 0.00 |
Nb FLOP rsqrt | 0.00 |
Bytes/cycle | 18.00 |
Bytes prefetched | 0.00 |
Bytes loaded | 68.00 |
Bytes stored | 16.00 |
Stride 0 | 1.00 |
Stride 1 | 1.00 |
Stride n | 1.00 |
Stride unknown | 2.00 |
Stride indirect | 2.00 |
Vectorization ratio all | 22.22 |
Vectorization ratio load | 0.00 |
Vectorization ratio store | 33.33 |
Vectorization ratio mul | NA |
Vectorization ratio add_sub | 50.00 |
Vectorization ratio fma | NA |
Vectorization ratio div_sqrt | NA |
Vectorization ratio other | 0.00 |
Vector-efficiency ratio all | 11.11 |
Vector-efficiency ratio load | 9.38 |
Vector-efficiency ratio store | 8.33 |
Vector-efficiency ratio mul | NA |
Vector-efficiency ratio add_sub | 15.63 |
Vector-efficiency ratio fma | NA |
Vector-efficiency ratio div_sqrt | NA |
Vector-efficiency ratio other | 12.50 |
Path / |
Function | dd_pmeredist_f(gmx_pme_t*, PmeAtomComm*, gmx::ArrayRef |
Source file and lines | pme_redistribute.cpp:436-449 |
Module | libgromacs_mpi.so.9.0.0 |
nb instructions | 20.50 |
nb uops | 19.50 |
loop length | 87.50 |
used x86 registers | 9 |
used mmx registers | 0 |
used xmm registers | 3 |
used ymm registers | 0 |
used zmm registers | 0 |
nb stack references | 1 |
micro-operation queue | 3.25 cycles |
front end | 3.25 cycles |
ALU0/BRU0 | ALU1 | ALU2 | ALU3 | BRU1 | AGU0 | AGU1 | AGU2 | FP0 | FP1 | FP2 | FP3 | FP4 | FP5 | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
uops | 2.13 | 2.13 | 2.00 | 2.00 | 1.25 | 3.83 | 3.83 | 3.83 | 0.00 | 0.00 | 1.00 | 1.00 | 1.00 | 1.00 |
cycles | 2.13 | 2.13 | 2.00 | 2.00 | 1.25 | 3.83 | 3.83 | 3.83 | 0.00 | 0.00 | 1.00 | 1.00 | 1.00 | 1.00 |
Cycles executing div or sqrt instructions | NA |
Longest recurrence chain latency (RecMII) | 2.00 |
Front-end | 3.25 |
Dispatch | 3.83 |
Data deps. | 2.00 |
Overall L1 | 3.83 |
all | 0% |
load | 0% |
store | 0% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | NA (no add-sub vectorizable/vectorized instructions) |
fma | NA (no fma vectorizable/vectorized instructions) |
other | 0% |
all | 28% |
load | 0% |
store | 50% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | 50% |
fma | NA (no fma vectorizable/vectorized instructions) |
div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
other | NA (no other vectorizable/vectorized instructions) |
all | 22% |
load | 0% |
store | 41% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | 50% |
fma | NA (no fma vectorizable/vectorized instructions) |
div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
other | 0% |
all | 9% |
load | 6% |
store | 6% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | NA (no add-sub vectorizable/vectorized instructions) |
fma | NA (no fma vectorizable/vectorized instructions) |
other | 10% |
all | 11% |
load | 9% |
store | 9% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | 15% |
fma | NA (no fma vectorizable/vectorized instructions) |
div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
other | NA (no other vectorizable/vectorized instructions) |
all | 11% |
load | 9% |
store | 8% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | 15% |
fma | NA (no fma vectorizable/vectorized instructions) |
div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
other | 10% |
Function | dd_pmeredist_f(gmx_pme_t*, PmeAtomComm*, gmx::ArrayRef |
Source file and lines | pme_redistribute.cpp:436-449 |
Module | libgromacs_mpi.so.9.0.0 |
nb instructions | 19 |
nb uops | 18 |
loop length | 79 |
used x86 registers | 8 |
used mmx registers | 0 |
used xmm registers | 3 |
used ymm registers | 0 |
used zmm registers | 0 |
nb stack references | 0 |
micro-operation queue | 3.00 cycles |
front end | 3.00 cycles |
ALU0/BRU0 | ALU1 | ALU2 | ALU3 | BRU1 | AGU0 | AGU1 | AGU2 | FP0 | FP1 | FP2 | FP3 | FP4 | FP5 | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
uops | 2.25 | 2.25 | 2.00 | 2.00 | 1.50 | 3.00 | 3.00 | 3.00 | 0.00 | 0.00 | 1.00 | 1.00 | 1.00 | 1.00 |
cycles | 2.25 | 2.25 | 2.00 | 2.00 | 1.50 | 3.00 | 3.00 | 3.00 | 0.00 | 0.00 | 1.00 | 1.00 | 1.00 | 1.00 |
Cycles executing div or sqrt instructions | NA |
Longest recurrence chain latency (RecMII) | 2.00 |
Front-end | 3.00 |
Dispatch | 3.00 |
Data deps. | 2.00 |
Overall L1 | 3.00 |
all | 0% |
load | 0% |
store | NA (no store vectorizable/vectorized instructions) |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | NA (no add-sub vectorizable/vectorized instructions) |
fma | NA (no fma vectorizable/vectorized instructions) |
other | 0% |
all | 28% |
load | 0% |
store | 50% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | 50% |
fma | NA (no fma vectorizable/vectorized instructions) |
div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
other | NA (no other vectorizable/vectorized instructions) |
all | 22% |
load | 0% |
store | 50% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | 50% |
fma | NA (no fma vectorizable/vectorized instructions) |
div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
other | 0% |
all | 9% |
load | 6% |
store | NA (no store vectorizable/vectorized instructions) |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | NA (no add-sub vectorizable/vectorized instructions) |
fma | NA (no fma vectorizable/vectorized instructions) |
other | 9% |
all | 11% |
load | 9% |
store | 9% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | 15% |
fma | NA (no fma vectorizable/vectorized instructions) |
div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
other | NA (no other vectorizable/vectorized instructions) |
all | 11% |
load | 8% |
store | 9% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | 15% |
fma | NA (no fma vectorizable/vectorized instructions) |
div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
other | 9% |
Instruction | Nb FU | ALU0/BRU0 | ALU1 | ALU2 | ALU3 | BRU1 | AGU0 | AGU1 | AGU2 | FP0 | FP1 | FP2 | FP3 | FP4 | FP5 | Latency | Recip. throughput | Vectorization |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
INC %RSI | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.25 | N/A |
ADD $0xc,%RBX | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.25 | N/A |
CMP %RSI,%RCX | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.25 | scal (12.5%) |
JE c0a469 <_Z14dd_pmeredist_fP9gmx_pme_tP11PmeAtomCommN3gmx8ArrayRefINS3_11BasicVectorIfEEEEb+0x369> | 1 | 0.50 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.50-1 | N/A |
MOVSXD (%RAX,%RSI,4),%RDI | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.33 | N/A |
CMP 0x8(%R10),%EDI | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.33 | scal (6.3%) |
JNE c0a300 <_Z14dd_pmeredist_fP9gmx_pme_tP11PmeAtomCommN3gmx8ArrayRefINS3_11BasicVectorIfEEEEb+0x200> | 1 | 0.50 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.50-1 | N/A |
MOVSXD %EDX,%RDX | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.25 | N/A |
MOV 0xa0(%R10),%RDI | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 3 | 0.33 | N/A |
LEA (%RDX,%RDX,2),%R8 | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.25 | N/A |
VMOVSS (%RBX),%XMM0 | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.50 | scal (6.3%) |
VADDSS 0x8(%RDI,%R8,4),%XMM0,%XMM0 | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0.50 | 0.50 | 0 | 0 | 3 | 0.50 | scal (6.3%) |
VMOVSD -0x8(%RBX),%XMM1 | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.50 | scal (12.5%) |
VMOVSD (%RDI,%R8,4),%XMM2 | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.50 | scal (12.5%) |
VADDPS %XMM2,%XMM1,%XMM1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0.50 | 0 | 0 | 3 | 0.50 | vect (25.0%) |
VMOVLPS %XMM1,-0x8(%RBX) | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0.50 | 0.50 | 1 | 1 | vect (12.5%) |
VMOVSS %XMM0,(%RBX) | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0.50 | 0.50 | 1 | 1 | scal (6.3%) |
INC %EDX | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.25 | N/A |
JMP c0a346 <_Z14dd_pmeredist_fP9gmx_pme_tP11PmeAtomCommN3gmx8ArrayRefINS3_11BasicVectorIfEEEEb+0x246> | 1 | 0.50 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | N/A |
Function | dd_pmeredist_f(gmx_pme_t*, PmeAtomComm*, gmx::ArrayRef |
Source file and lines | pme_redistribute.cpp:436-449 |
Module | libgromacs_mpi.so.9.0.0 |
nb instructions | 22 |
nb uops | 21 |
loop length | 96 |
used x86 registers | 10 |
used mmx registers | 0 |
used xmm registers | 3 |
used ymm registers | 0 |
used zmm registers | 0 |
nb stack references | 2 |
micro-operation queue | 3.50 cycles |
front end | 3.50 cycles |
ALU0/BRU0 | ALU1 | ALU2 | ALU3 | BRU1 | AGU0 | AGU1 | AGU2 | FP0 | FP1 | FP2 | FP3 | FP4 | FP5 | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
uops | 2.00 | 2.00 | 2.00 | 2.00 | 1.00 | 4.67 | 4.67 | 4.67 | 0.00 | 0.00 | 1.00 | 1.00 | 1.00 | 1.00 |
cycles | 2.00 | 2.00 | 2.00 | 2.00 | 1.00 | 4.67 | 4.67 | 4.67 | 0.00 | 0.00 | 1.00 | 1.00 | 1.00 | 1.00 |
Cycles executing div or sqrt instructions | NA |
Longest recurrence chain latency (RecMII) | 2.00 |
Front-end | 3.50 |
Dispatch | 4.67 |
Data deps. | 2.00 |
Overall L1 | 4.67 |
all | 0% |
load | NA (no load vectorizable/vectorized instructions) |
store | 0% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | NA (no add-sub vectorizable/vectorized instructions) |
fma | NA (no fma vectorizable/vectorized instructions) |
other | 0% |
all | 28% |
load | 0% |
store | 50% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | 50% |
fma | NA (no fma vectorizable/vectorized instructions) |
div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
other | NA (no other vectorizable/vectorized instructions) |
all | 22% |
load | 0% |
store | 33% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | 50% |
fma | NA (no fma vectorizable/vectorized instructions) |
div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
other | 0% |
all | 9% |
load | NA (no load vectorizable/vectorized instructions) |
store | 6% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | NA (no add-sub vectorizable/vectorized instructions) |
fma | NA (no fma vectorizable/vectorized instructions) |
other | 12% |
all | 11% |
load | 9% |
store | 9% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | 15% |
fma | NA (no fma vectorizable/vectorized instructions) |
div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
other | NA (no other vectorizable/vectorized instructions) |
all | 11% |
load | 9% |
store | 8% |
mul | NA (no mul vectorizable/vectorized instructions) |
add-sub | 15% |
fma | NA (no fma vectorizable/vectorized instructions) |
div/sqrt | NA (no div/sqrt vectorizable/vectorized instructions) |
other | 12% |
Instruction | Nb FU | ALU0/BRU0 | ALU1 | ALU2 | ALU3 | BRU1 | AGU0 | AGU1 | AGU2 | FP0 | FP1 | FP2 | FP3 | FP4 | FP5 | Latency | Recip. throughput | Vectorization |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
MOV -0x50(%RBP),%R8 | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 3 | 0.33 | N/A |
MOV 0x768(%R8),%R8 | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 3 | 0.33 | N/A |
MOV 0x30(%R10),%R9 | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 3 | 0.33 | N/A |
MOVSXD (%R9,%RDI,4),%R10 | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.33 | N/A |
LEA (%R10,%R10,2),%R11 | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.25 | N/A |
VMOVSS (%RBX),%XMM0 | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.50 | scal (6.3%) |
VADDSS 0x8(%R8,%R11,4),%XMM0,%XMM0 | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0.50 | 0.50 | 0 | 0 | 3 | 0.50 | scal (6.3%) |
VMOVSD -0x8(%RBX),%XMM1 | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.50 | scal (12.5%) |
VMOVSD (%R8,%R11,4),%XMM2 | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.50 | scal (12.5%) |
VADDPS %XMM2,%XMM1,%XMM1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.50 | 0.50 | 0 | 0 | 3 | 0.50 | vect (25.0%) |
VMOVLPS %XMM1,-0x8(%RBX) | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0.50 | 0.50 | 1 | 1 | vect (12.5%) |
VMOVSS %XMM0,(%RBX) | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0.50 | 0.50 | 1 | 1 | scal (6.3%) |
LEA 0x1(%R10),%R8D | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.25 | N/A |
MOV -0x30(%RBP),%R10 | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 3 | 0.33 | N/A |
MOV %R8D,(%R9,%RDI,4) | 1 | 0 | 0 | 0 | 0 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 0.50 | scal (6.3%) |
INC %RSI | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.25 | N/A |
ADD $0xc,%RBX | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.25 | N/A |
CMP %RSI,%RCX | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.25 | scal (12.5%) |
JE c0a469 <_Z14dd_pmeredist_fP9gmx_pme_tP11PmeAtomCommN3gmx8ArrayRefINS3_11BasicVectorIfEEEEb+0x369> | 1 | 0.50 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.50-1 | N/A |
MOVSXD (%RAX,%RSI,4),%RDI | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.33 | N/A |
CMP 0x8(%R10),%EDI | 1 | 0.25 | 0.25 | 0.25 | 0.25 | 0 | 0.33 | 0.33 | 0.33 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.33 | N/A |
JNE c0a300 <_Z14dd_pmeredist_fP9gmx_pme_tP11PmeAtomCommN3gmx8ArrayRefINS3_11BasicVectorIfEEEEb+0x200> | 1 | 0.50 | 0 | 0 | 0 | 0.50 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0.50-1 | N/A |
Run 1x1 | Number processes: 1Number processes per node: 1OMP_NUM_THREADS: 1 |
---|---|
Run 2x1 | Number processes: 2Number processes per node: 2OMP_NUM_THREADS: 1 |
Run 4x1 | Number processes: 4Number processes per node: 4OMP_NUM_THREADS: 1 |
Run 8x1 | Number processes: 8Number processes per node: 8OMP_NUM_THREADS: 1 |
Run 16x1 | Number processes: 16Number processes per node: 16OMP_NUM_THREADS: 1 |
Run 32x1 | Number processes: 32Number processes per node: 32OMP_NUM_THREADS: 1 |
Run 64x1 | Number processes: 64Number processes per node: 64OMP_NUM_THREADS: 1 |
Run 128x1 | Number processes: 128Number processes per node: 128OMP_NUM_THREADS: 1 |
Run 192x1 | Number processes: 192Number nodes: 1Number processes per node: 192Run Command: <executable> mdrun -s ion_channel.tpr -nsteps 10000 -pin on -deffnm aoccMPI Command: mpirun -genv I_MPI_FABRICS=shm -n <number_processes>Dataset: Run Directory: .OMP_NUM_THREADS: 1 |
(1x1) Efficiency | (1x1) Potential Speed-Up (%) | (2x1) Efficiency | (2x1) Potential Speed-Up (%) | (4x1) Efficiency | (4x1) Potential Speed-Up (%) | (8x1) Efficiency | (8x1) Potential Speed-Up (%) | (16x1) Efficiency | (16x1) Potential Speed-Up (%) | (32x1) Efficiency | (32x1) Potential Speed-Up (%) | (64x1) Efficiency | (64x1) Potential Speed-Up (%) | (128x1) Efficiency | (128x1) Potential Speed-Up (%) | (192x1) Efficiency | (192x1) Potential Speed-Up (%) |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | 0 | 1 | 0 | 1 | 0 | 1 | 0 |
Run | Number of threads | Efficiency (ideal is 1) | Speedup | Ideal Speedup | Time (s) | Coverage (%) |
---|---|---|---|---|---|---|
1x1 | ||||||
2x1 | 2 | 1 | 1 | 2 | 0.92999988794327 | 0.17925477027893 |
4x1 | 4 | 1 | 1 | 4 | 0.63499999046326 | 0.21488237380981 |
8x1 | 8 | 1 | 1 | 8 | 0.59999990463257 | 0.29337507486343 |
16x1 | 16 | 1 | 1 | 16 | 0.27500003576279 | 0.24518243968487 |
32x1 | ||||||
64x1 | ||||||
128x1 | ||||||
192x1 |