OV - exec - Function qmcplusplus::DistanceTableBA<double, 3u, 39>::evaluate(qmcplusplus::ParticleSet&)

	Function: qmcplusplus::DistanceTableBA<double, 3u, 39>::evaluate(qmcplusplus::ParticleSet&)	Module: exec	Source: DistanceTableBA.h:65-105 [...]	Coverage: 0.1%

/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Particle/Lattice/ParticleBConds.h: 235 - 278

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

235:     const T* restrict py = R0 + r0_stride;

236:     const T* restrict pz = R0 + r0_stride * 2;

[...]

249:     const T flip    = iat < flip_ind ? one : minusone;

250:     const T displ_0 = (px[iat] - x0) * flip;

251:     const T displ_1 = (py[iat] - y0) * flip;

252:     const T displ_2 = (pz[iat] - z0) * flip;

253:

254:     const T ar_0 = -std::floor(displ_0 * g00 + displ_1 * g10 + displ_2 * g20);

255:     const T ar_1 = -std::floor(displ_0 * g01 + displ_1 * g11 + displ_2 * g21);

256:     const T ar_2 = -std::floor(displ_0 * g02 + displ_1 * g12 + displ_2 * g22);

257:

258:     const T delx = displ_0 + ar_0 * r00 + ar_1 * r10 + ar_2 * r20;

259:     const T dely = displ_1 + ar_0 * r01 + ar_1 * r11 + ar_2 * r21;

260:     const T delz = displ_2 + ar_0 * r02 + ar_1 * r12 + ar_2 * r22;

261:

262:     T rmin = delx * delx + dely * dely + delz * delz;

263:     int ic = 0;

264: #pragma unroll(7)

265:     for (int c = 1; c < 8; ++c)

266:     {

267:       const T x  = delx + cellx[c];

268:       const T y  = dely + celly[c];

269:       const T z  = delz + cellz[c];

270:       const T r2 = x * x + y * y + z * z;

271:       ic         = (r2 < rmin) ? c : ic;

272:       rmin       = (r2 < rmin) ? r2 : rmin;

273:     }

274:

275:     temp_r[iat] = std::sqrt(rmin);

276:     dx[iat]     = flip * (delx + cellx[ic]);

277:     dy[iat]     = flip * (dely + celly[ic]);

278:     dz[iat]     = flip * (delz + cellz[ic]);

/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Particle/DistanceTableBA.h: 65 - 105

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

65:   inline void evaluate(ParticleSet& P)

[...]

73:     const size_t ntgt_local  = Ntargets;

[...]

88:     for (size_t iat = 0; iat < ntgt_local; ++iat)

89:       for (size_t team_id = 0; team_id < num_teams; ++team_id)

90:       {

91:         T pos[D];

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

93:           pos[idim] = *(tgt_ptr + ntgt_padded * idim + iat);

94:

95:         const size_t first = ChunkSizePerTeam * team_id;

96:         const size_t last  = std::min(first + ChunkSizePerTeam, nsrc_padded);

97:

98:         PRAGMA_OFFLOAD("omp parallel for")

99:         for (size_t jel = first; jel < last; ++jel)

100:         {

101:           DTD_BConds<T, D, SC>::computeDistancesOffload(pos, src_ptr, nsrc_padded, dist_ptr + nsrc_padded * iat,

102:                                                         displ_ptr + nsrc_padded * D * iat, nsrc_padded, jel);

103:         }

104:       }

105:   }

/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsVector.h: 248 - 248

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

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

/usr/include/c++/11/bits/stl_vector.h: 1169 - 1169

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

1169:       { return _M_data_ptr(this->_M_impl._M_start); }

/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/VectorSoAContainer.h: 237 - 239

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

237:   T* data() { return myData; }

238:   ///return the base

239:   const T* data() const { return myData; }

0x4700a0 STP	X29, X30, [SP, #656]!

0x4700a4 ADD	X29, SP, #0

0x4700a8 ORR	X8, XZR, X0

0x4700ac STP	X21, X22, [SP, #32]

0x4700b0 STP	X27, X28, [SP, #80]

0x4700b4 LDPSW	W21, W28, [X0, #196]

0x4700b8 STP	X23, X24, [SP, #48]

0x4700bc LDR	X0, [X0, #8]

0x4700c0 LDR	X30, [X8, #64]

0x4700c4 LDR	X3, [X8, #168]

0x4700c8 LDR	X4, [X0, #1552]

0x4700cc LDR	X24, [X1, #1552]

0x4700d0 CBZ	X28, 470510

0x4700d4 ADD	X22, X28, #3

0x4700d8 ADD	X1, X21, #3

0x4700dc STP	X25, X26, [SP, #64]

0x4700e0 AND	X21, X1, #8062

0x4700e4 AND	X2, X22, #8062

0x4700e8 UBFM	X5, X2, #61, #60

0x4700ec ADD	X26, X21, X21,LSL #1

0x4700f0 ADD	X27, X21, #127

0x4700f4 UBFM	X27, X27, #7, #63

0x4700f8 ADD	X23, X24, X5

0x4700fc STP	X19, X20, [SP, #16]

0x470100 UBFM	X6, X21, #61, #60

0x470104 SBFM	X18, X21, #63, #30

0x470108 ADD	X22, X23, X5

0x47010c UBFM	X26, X26, #61, #60

0x470110 UBFM	X25, X27, #57, #56

0x470114 MOVZ	X10, #0

0x470118 STP	D8, D9, [SP, #96]

0x47011c MOVZ	X20, #0

0x470120 SBFM	X17, X21, #0, #31

0x470124 STP	D10, D11, [SP, #112]

0x470128 MOVZ	W16, #2

0x47012c MOVZ	W15, #3

0x470130 STP	D12, D13, [SP, #128]

0x470134 MOVZ	W14, #4

0x470138 MOVZ	W13, #5

0x47013c STP	D14, D15, [SP, #144]

0x470140 MOVZ	W12, #6

0x470144 MOVZ	W11, #7

0x470148 STR	X6, [SP, #360]

(1240) 0x47014c CBZ	X27, 4704e0

0x470150 MOVZ	X19, #0

0x470154 HINT	#0

0x470158 HINT	#0

0x47015c HINT	#0

(1239) 0x470160 ORR	X1, XZR, X19

(1239) 0x470164 ADD	X19, X19, #128

(1239) 0x470168 LDR	D16, [X24, X20,LSL #3]

(1239) 0x47016c CMP	X19, X21

(1239) 0x470170 CSEL	X9, X19, X21, #9

(1239) 0x470174 LDR	D15, [X23, X20,LSL #3]

(1239) 0x470178 LDR	D14, [X22, X20,LSL #3]

(1239) 0x47017c CMP	X9, X1

(1239) 0x470180 B.LS	4704d8

(1239) 0x470184 ADD	X7, X8, #240

(1239) 0x470188 LDR	D3, [X7, #40]

(1239) 0x47018c LDR	D6, [X7, #48]

(1239) 0x470190 LDR	D7, [X7, #56]

(1239) 0x470194 LDR	D8, [X7, #64]

(1239) 0x470198 STR	D3, [SP, #160]

(1239) 0x47019c LDP	D20, D0, [X8, #408]

(1239) 0x4701a0 STR	D6, [SP, #168]

(1239) 0x4701a4 LDR	D4, [X8, #424]

(1239) 0x4701a8 STR	D7, [SP, #176]

(1239) 0x4701ac LDR	D1, [X8, #480]

(1239) 0x4701b0 STR	D8, [SP, #184]

(1239) 0x4701b4 LDR	D5, [X8, #488]

(1239) 0x4701b8 STR	D0, [SP, #352]

(1239) 0x4701bc LDR	D2, [X8, #544]

(1239) 0x4701c0 STR	D4, [SP, #328]

(1239) 0x4701c4 STR	D1, [SP, #344]

(1239) 0x4701c8 LDP	D29, D25, [X7, #8]

(1239) 0x4701cc STR	D5, [SP, #320]

(1239) 0x4701d0 LDP	D21, D17, [X7, #24]

(1239) 0x4701d4 STR	D2, [SP, #336]

(1239) 0x4701d8 LDR	D30, [X8, #240]

(1239) 0x4701dc LDP	D28, D24, [X8, #392]

(1239) 0x4701e0 LDP	D27, D23, [X8, #456]

(1239) 0x4701e4 LDR	D19, [X8, #472]

(1239) 0x4701e8 LDR	D26, [X8, #520]

(1239) 0x4701ec LDR	D22, [X8, #528]

(1239) 0x4701f0 LDR	D18, [X8, #536]

(1239) 0x4701f4 LDR	D9, [X7, #72]

(1239) 0x4701f8 LDR	D10, [X7, #80]

(1239) 0x4701fc LDR	D11, [X7, #88]

(1239) 0x470200 STR	D9, [SP, #192]

(1239) 0x470204 LDR	D12, [X7, #96]

(1239) 0x470208 STR	D10, [SP, #200]

(1239) 0x47020c LDR	D13, [X7, #104]

(1239) 0x470210 STR	D11, [SP, #208]

(1239) 0x470214 LDR	D31, [X7, #112]

(1239) 0x470218 LDR	D0, [X7, #120]

(1239) 0x47021c STR	D12, [SP, #216]

(1239) 0x470220 LDR	D1, [X7, #128]

(1239) 0x470224 STR	D13, [SP, #224]

(1239) 0x470228 LDR	D2, [X7, #136]

(1239) 0x47022c STR	D31, [SP, #232]

(1239) 0x470230 LDR	D4, [X8, #432]

(1239) 0x470234 STR	D0, [SP, #240]

(1239) 0x470238 LDR	D7, [X8, #440]

(1239) 0x47023c STR	D1, [SP, #248]

(1239) 0x470240 LDR	D5, [X8, #496]

(1239) 0x470244 STR	D2, [SP, #256]

(1239) 0x470248 LDR	D8, [X8, #504]

(1239) 0x47024c STR	D4, [SP, #304]

(1239) 0x470250 LDR	D3, [X8, #552]

(1239) 0x470254 STR	D7, [SP, #264]

(1239) 0x470258 LDR	D6, [X8, #560]

(1239) 0x47025c STR	D5, [SP, #296]

(1239) 0x470260 LDR	D9, [X8, #568]

(1239) 0x470264 STR	D8, [SP, #272]

(1239) 0x470268 STR	D3, [SP, #312]

(1239) 0x47026c STR	D6, [SP, #288]

(1239) 0x470270 STR	D9, [SP, #280]

(1239) 0x470274 HINT	#0

(1239) 0x470278 HINT	#0

(1239) 0x47027c HINT	#0

  (1241) 0x470280 SBFM	X2, X1, #0, #31

  (1241) 0x470284 FMOV	D11, #-1.0000000

  (1241) 0x470288 FMOV	D12, #1.0000000

  (1241) 0x47028c ADD	X6, X2, X17

  (1241) 0x470290 CMP	W1, #0

  (1241) 0x470294 ADD	X7, X30, W1,SXTW #3

  (1241) 0x470298 LDR	D10, [X4, X2,LSL #3]

  (1241) 0x47029c ADD	X5, X18, X2

  (1241) 0x4702a0 ADD	X1, X1, #1

  (1241) 0x4702a4 LDR	D13, [X4, X6,LSL #3]

  (1241) 0x4702a8 FCSEL	D5, D11, D12, #10

  (1241) 0x4702ac LDR	D31, [X4, X5,LSL #3]

  (1241) 0x4702b0 FSUB	D0, D10, S16

  (1241) 0x4702b4 FSUB	D1, D13, S15

  (1241) 0x4702b8 LDR	D13, [SP, #176]

  (1241) 0x4702bc FSUB	D2, D31, S14

  (1241) 0x4702c0 LDR	D31, [SP, #200]

  (1241) 0x4702c4 FMUL	D6, D0, D5

  (1241) 0x4702c8 FMUL	D8, D1, D5

  (1241) 0x4702cc LDR	D1, [SP, #168]

  (1241) 0x4702d0 FMUL	D7, D2, D5

  (1241) 0x4702d4 LDR	D2, [SP, #160]

  (1241) 0x4702d8 FMUL	D4, D8, D29

  (1241) 0x4702dc FMUL	D3, D8, D17

  (1241) 0x4702e0 FMUL	D0, D8, D13

  (1241) 0x4702e4 LDR	D13, [SP, #224]

  (1241) 0x4702e8 FMADD	D9, D6, D30, D4

  (1241) 0x4702ec FMADD	D10, D6, D21, D3

  (1241) 0x4702f0 LDR	D4, [SP, #248]

  (1241) 0x4702f4 FMADD	D3, D6, D1, D0

  (1241) 0x4702f8 LDR	D0, [SP, #184]

  (1241) 0x4702fc FMADD	D11, D7, D25, D9

  (1241) 0x470300 FMADD	D10, D7, D2, D10

  (1241) 0x470304 LDR	D2, [SP, #216]

  (1241) 0x470308 FRINTM	D12, D11

  (1241) 0x47030c FMADD	D1, D7, D0, D3

  (1241) 0x470310 FRINTM	D9, D10

  (1241) 0x470314 FNMUL	D11, D9, D31

  (1241) 0x470318 FNMUL	D31, D9, D13

  (1241) 0x47031c FNMUL	D4, D9, D4

  (1241) 0x470320 LDR	D9, [SP, #192]

  (1241) 0x470324 FRINTM	D3, D1

  (1241) 0x470328 FMSUB	D10, D12, D2, D31

  (1241) 0x47032c LDR	D31, [SP, #240]

  (1241) 0x470330 FMSUB	D11, D12, D9, D11

  (1241) 0x470334 LDR	D9, [SP, #208]

  (1241) 0x470338 FMSUB	D12, D12, D31, D4

  (1241) 0x47033c LDR	D4, [SP, #272]

  (1241) 0x470340 FADD	D8, D10, D8

  (1241) 0x470344 FADD	D13, D6, D11

  (1241) 0x470348 LDR	D6, [SP, #344]

  (1241) 0x47034c LDR	D11, [SP, #320]

  (1241) 0x470350 FADD	D0, D12, D7

  (1241) 0x470354 LDR	D7, [SP, #232]

  (1241) 0x470358 FMSUB	D1, D3, D7, D8

  (1241) 0x47035c LDR	D7, [SP, #328]

  (1241) 0x470360 FADD	D2, D1, D19

  (1241) 0x470364 FADD	D8, D1, D4

  (1241) 0x470368 LDR	D4, [SP, #256]

  (1241) 0x47036c FADD	D10, D1, D6

  (1241) 0x470370 FMUL	D12, D2, D2

  (1241) 0x470374 FMSUB	D2, D3, D9, D13

  (1241) 0x470378 FMSUB	D0, D3, D4, D0

  (1241) 0x47037c FADD	D4, D1, D27

  (1241) 0x470380 FMUL	D31, D10, D10

  (1241) 0x470384 FMUL	D10, D8, D8

  (1241) 0x470388 FADD	D3, D1, D23

  (1241) 0x47038c FADD	D13, D1, D11

  (1241) 0x470390 FMUL	D4, D4, D4

  (1241) 0x470394 FMUL	D3, D3, D3

  (1241) 0x470398 FADD	D6, D2, D20

  (1241) 0x47039c FADD	D8, D2, D28

  (1241) 0x4703a0 FADD	D11, D2, D24

  (1241) 0x4703a4 FADD	D7, D2, D7

  (1241) 0x4703a8 FMUL	D13, D13, D13

  (1241) 0x4703ac FMADD	D9, D6, D6, D12

  (1241) 0x4703b0 FMUL	D6, D1, D1

  (1241) 0x4703b4 LDR	D12, [SP, #296]

  (1241) 0x4703b8 FMADD	D8, D8, D8, D4

  (1241) 0x4703bc FADD	D4, D0, D26

  (1241) 0x4703c0 FMADD	D11, D11, D11, D3

  (1241) 0x4703c4 FMADD	D13, D7, D7, D13

  (1241) 0x4703c8 LDR	D7, [SP, #304]

  (1241) 0x4703cc FMADD	D6, D2, D2, D6

  (1241) 0x4703d0 FADD	D12, D1, D12

  (1241) 0x4703d4 FADD	D7, D2, D7

  (1241) 0x4703d8 FMADD	D3, D0, D0, D6

  (1241) 0x4703dc FMADD	D6, D4, D4, D8

  (1241) 0x4703e0 FMINNM	D8, D3, D6

  (1241) 0x4703e4 FADD	D4, D0, D18

  (1241) 0x4703e8 FMUL	D12, D12, D12

  (1241) 0x4703ec FMADD	D12, D7, D7, D12

  (1241) 0x4703f0 LDR	D7, [SP, #264]

  (1241) 0x4703f4 FCMPE	D3, D6

  (1241) 0x4703f8 LDR	D3, [SP, #352]

  (1241) 0x4703fc FMADD	D6, D4, D4, D9

  (1241) 0x470400 CSINC	W0, WZR, WZR, #13

  (1241) 0x470404 LDR	D9, [SP, #288]

  (1241) 0x470408 FADD	D4, D2, D3

  (1241) 0x47040c FADD	D3, D0, D22

  (1241) 0x470410 FADD	D9, D0, D9

  (1241) 0x470414 FMADD	D11, D3, D3, D11

  (1241) 0x470418 LDR	D3, [SP, #336]

  (1241) 0x47041c FMADD	D31, D4, D4, D31

  (1241) 0x470420 FMADD	D4, D9, D9, D12

  (1241) 0x470424 FADD	D12, D2, D7

  (1241) 0x470428 FMADD	D9, D12, D12, D10

  (1241) 0x47042c FCMPE	D11, D8

  (1241) 0x470430 FMINNM	D8, D11, D8

  (1241) 0x470434 FADD	D11, D0, D3

  (1241) 0x470438 CSEL	W0, W0, W16, #10

  (1241) 0x47043c FCMPE	D6, D8

  (1241) 0x470440 FMINNM	D6, D6, D8

  (1241) 0x470444 LDR	D8, [SP, #312]

  (1241) 0x470448 CSEL	W0, W0, W15, #10

  (1241) 0x47044c FMADD	D10, D11, D11, D31

  (1241) 0x470450 FMINNM	D31, D10, D6

  (1241) 0x470454 FCMPE	D10, D6

  (1241) 0x470458 FADD	D6, D0, D8

  (1241) 0x47045c CSEL	W0, W0, W14, #10

  (1241) 0x470460 FMADD	D13, D6, D6, D13

  (1241) 0x470464 FMINNM	D7, D13, D31

  (1241) 0x470468 FMINNM	D3, D4, D7

  (1241) 0x47046c FCMPE	D13, D31

  (1241) 0x470470 CSEL	W0, W0, W13, #10

  (1241) 0x470474 FCMPE	D4, D7

  (1241) 0x470478 LDR	D4, [SP, #280]

  (1241) 0x47047c CSEL	W0, W0, W12, #10

  (1241) 0x470480 FADD	D12, D0, D4

  (1241) 0x470484 FMADD	D11, D12, D12, D9

  (1241) 0x470488 FMINNM	D9, D11, D3

  (1241) 0x47048c FSQRT	D10, D9

  (1241) 0x470490 FCMPE	D11, D3

  (1241) 0x470494 CSEL	W0, W0, W11, #10

  (1241) 0x470498 ADD	X0, X8, W0,SXTW #3

  (1241) 0x47049c LDR	D31, [X0, #384]

  (1241) 0x4704a0 LDR	D8, [X0, #448]

  (1241) 0x4704a4 LDR	D6, [X0, #512]

  (1241) 0x4704a8 STR	D10, [X7, X10]

  (1241) 0x4704ac FADD	D2, D2, D31

  (1241) 0x4704b0 FADD	D1, D1, D8

  (1241) 0x4704b4 FADD	D0, D0, D6

  (1241) 0x4704b8 FMUL	D13, D2, D5

  (1241) 0x4704bc FMUL	D7, D1, D5

  (1241) 0x4704c0 FMUL	D5, D0, D5

  (1241) 0x4704c4 STR	D13, [X3, X2,LSL #3]

  (1241) 0x4704c8 STR	D7, [X3, X6,LSL #3]

  (1241) 0x4704cc STR	D5, [X3, X5,LSL #3]

  (1241) 0x4704d0 CMP	X1, X9

  (1241) 0x4704d4 B.NE	470280

(1239) 0x4704d8 CMP	X19, X25

(1239) 0x4704dc B.NE	470160

(1240) 0x4704e0 LDR	X19, [SP, #360]

(1240) 0x4704e4 ADD	X20, X20, #1

(1240) 0x4704e8 ADD	X3, X3, X26

(1240) 0x4704ec ADD	X10, X10, X19

(1240) 0x4704f0 CMP	X28, X20

(1240) 0x4704f4 B.NE	47014c

0x4704f8 LDP	D8, D9, [SP, #96]

0x4704fc LDP	X19, X20, [SP, #16]

0x470500 LDP	X25, X26, [SP, #64]

0x470504 LDP	D10, D11, [SP, #112]

0x470508 LDP	D12, D13, [SP, #128]

0x47050c LDP	D14, D15, [SP, #144]

0x470510 LDP	X21, X22, [SP, #32]

0x470514 LDP	X23, X24, [SP, #48]

0x470518 LDP	X27, X28, [SP, #80]

0x47051c LDP	X29, X30, [SP], #368

0x470520 RET

0x470524 HINT	#0

0x470528 HINT	#0

0x47052c HINT	#0

0x470530 HINT	#0

0x470534 HINT	#0

0x470538 HINT	#0

0x47053c HINT	#0

Coverage (%)	Name	Source Location	Module
►100.00+	qmcplusplus::ParticleSet::upda[...]	stl_vector.h:918	exec
○	main._omp_fn.0	miniqmc.cpp:390	exec
○	GOMP_parallel		libomp.so
○	main	stl_vector.h:1043	exec
○	__libc_start_main		libc-2.31.so
○	_start	miniqmc.cpp:546	exec

Path /

Average path: Display a virtual path defined by average values of all real paths

Source file and lines	DistanceTableBA.h:65-105
Module	exec

The function is defined in:

/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Particle/Lattice/ParticleBConds.h:235-236,249-278
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Particle/DistanceTableBA.h:65,73,88-105
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsVector.h:248
/usr/include/c++/11/bits/stl_vector.h:1169
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/VectorSoAContainer.h:237-239

Warnings:
Ignoring paths for analysis

gain
potential
hint
expert

Code clean check

Detected a slowdown caused by scalar integer instructions (typically used for address computation). By removing them, you can lower the cost of an iteration from 9.17 to 3.33 cycles (2.75x speedup).

Workaround

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

Vectorization

Your function is poorly vectorized. Only 30% of vector register length is used (average across all VPU instructions). By fully vectorizing your function, you can lower the cost of an iteration from 9.17 to 2.15 cycles (4.27x speedup).

Details

36% of VPU instructions are used in vector version (process two or more data elements in vector registers):

0% of VPU addition or subtraction instructions are used in vector version.
0% of VPU 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)

No data for this section

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 208 bytes. The binary function is storing 168 bytes.

General properties

nb instructions	65
loop length	260
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

6.88 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	1.00	1.00	7.00	7.00	7.00	7.00	2.00	2.00	0.00	0.00	9.17	8.83	9.00	3.50	3.50
cycles	1.00	1.00	7.00	7.00	7.00	7.00	2.00	2.00	0.00	0.00	9.17	8.83	9.00	3.50	3.50

Cycles executing div or sqrt instructions

Cycles summary

Front-end	6.88
Overall L1	9.17

Vectorization ratios

all	36%
load	100%
store	100%
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%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 9.17 cycles. At this rate:

23% of peak load performance is reached (22.69 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
38% of peak store performance is reached (18.33 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the function is: 4700a0

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput
STP X29, X30, [SP, #656]!	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
ADD X29, SP, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ORR X8, XZR, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP X21, X22, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X27, X28, [SP, #80]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
LDPSW W21, W28, [X0, #196]	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	5	1
STP X23, X24, [SP, #48]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
LDR X0, [X0, #8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X30, [X8, #64]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X3, [X8, #168]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X4, [X0, #1552]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X24, [X1, #1552]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CBZ X28, 470510 <_ZN11qmcplusplus15DistanceTableBAIdLj3ELi39EE8evaluateERNS_11ParticleSetE+0x470>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADD X22, X28, #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X1, X21, #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP X25, X26, [SP, #64]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
AND X21, X1, #8062	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
AND X2, X22, #8062	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
UBFM X5, X2, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X26, X21, X21,LSL #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X27, X21, #127	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
UBFM X27, X27, #7, #63	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X23, X24, X5	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP X19, X20, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
UBFM X6, X21, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SBFM X18, X21, #63, #30	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X22, X23, X5	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
UBFM X26, X26, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
UBFM X25, X27, #57, #56	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ X10, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP D8, D9, [SP, #96]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
MOVZ X20, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SBFM X17, X21, #0, #31	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP D10, D11, [SP, #112]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
MOVZ W16, #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ W15, #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP D12, D13, [SP, #128]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
MOVZ W14, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ W13, #5	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP D14, D15, [SP, #144]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
MOVZ W12, #6	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ W11, #7	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STR X6, [SP, #360]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
MOVZ X19, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
HINT #0
HINT #0
HINT #0
LDP D8, D9, [SP, #96]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDP X19, X20, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X25, X26, [SP, #64]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP D10, D11, [SP, #112]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDP D12, D13, [SP, #128]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDP D14, D15, [SP, #144]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDP X21, X22, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X23, X24, [SP, #48]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X27, X28, [SP, #80]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X29, X30, [SP], #368	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
RET	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0

Source file and lines	DistanceTableBA.h:65-105
Module	exec

The function is defined in:

/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Particle/Lattice/ParticleBConds.h:235-236,249-278
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Particle/DistanceTableBA.h:65,73,88-105
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/OhmmsVector.h:248
/usr/include/c++/11/bits/stl_vector.h:1169
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Numerics/OhmmsPETE/VectorSoAContainer.h:237-239

Warnings:
Ignoring paths for analysis

gain
potential
hint
expert

Code clean check

Detected a slowdown caused by scalar integer instructions (typically used for address computation). By removing them, you can lower the cost of an iteration from 9.17 to 3.33 cycles (2.75x speedup).

Workaround

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

Vectorization

Details

36% of VPU instructions are used in vector version (process two or more data elements in vector registers):

0% of VPU addition or subtraction instructions are used in vector version.
0% of VPU 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)

No data for this section

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 208 bytes. The binary function is storing 168 bytes.

General properties

nb instructions	65
loop length	260
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

6.88 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	1.00	1.00	7.00	7.00	7.00	7.00	2.00	2.00	0.00	0.00	9.17	8.83	9.00	3.50	3.50
cycles	1.00	1.00	7.00	7.00	7.00	7.00	2.00	2.00	0.00	0.00	9.17	8.83	9.00	3.50	3.50

Cycles executing div or sqrt instructions

Cycles summary

Front-end	6.88
Overall L1	9.17

Vectorization ratios

all	36%
load	100%
store	100%
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%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 9.17 cycles. At this rate:

23% of peak load performance is reached (22.69 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
38% of peak store performance is reached (18.33 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the function is: 4700a0

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput
STP X29, X30, [SP, #656]!	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
ADD X29, SP, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ORR X8, XZR, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP X21, X22, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X27, X28, [SP, #80]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
LDPSW W21, W28, [X0, #196]	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	5	1
STP X23, X24, [SP, #48]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
LDR X0, [X0, #8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X30, [X8, #64]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X3, [X8, #168]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X4, [X0, #1552]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X24, [X1, #1552]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CBZ X28, 470510 <_ZN11qmcplusplus15DistanceTableBAIdLj3ELi39EE8evaluateERNS_11ParticleSetE+0x470>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADD X22, X28, #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X1, X21, #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP X25, X26, [SP, #64]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
AND X21, X1, #8062	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
AND X2, X22, #8062	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
UBFM X5, X2, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X26, X21, X21,LSL #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X27, X21, #127	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
UBFM X27, X27, #7, #63	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X23, X24, X5	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP X19, X20, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
UBFM X6, X21, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SBFM X18, X21, #63, #30	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X22, X23, X5	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
UBFM X26, X26, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
UBFM X25, X27, #57, #56	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ X10, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP D8, D9, [SP, #96]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
MOVZ X20, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SBFM X17, X21, #0, #31	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP D10, D11, [SP, #112]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
MOVZ W16, #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ W15, #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP D12, D13, [SP, #128]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
MOVZ W14, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ W13, #5	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP D14, D15, [SP, #144]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
MOVZ W12, #6	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ W11, #7	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STR X6, [SP, #360]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
MOVZ X19, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
HINT #0
HINT #0
HINT #0
LDP D8, D9, [SP, #96]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDP X19, X20, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X25, X26, [SP, #64]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP D10, D11, [SP, #112]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDP D12, D13, [SP, #128]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDP D14, D15, [SP, #144]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDP X21, X22, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X23, X24, [SP, #48]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X27, X28, [SP, #80]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X29, X30, [SP], #368	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
RET	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0
HINT #0

Name	Coverage (%)	Time (s)
▼qmcplusplus::DistanceTableBA::evaluate(qmcplusplus::ParticleSet&)–	0.1	0.06
○Loop 1240 - DistanceTableBA.h:88-89 - exec	0	0
▼Loop 1239 - DistanceTableBA.h:89-99 - exec–	0	0
○Loop 1241 - DistanceTableBA.h:99-99 - exec	0.1	0.06

Report Configuration

Code clean check

Workaround

Vectorization

Details

Workaround

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

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Cycles and memory resources usage

ASM code

Code clean check

Workaround

Vectorization

Details

Workaround

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

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Cycles and memory resources usage

ASM code