Ticket #27831

Speed up Mndo::RotateTwoElecTwoCoreDiatomicToSpaceFramegc

Open Date: 2012-03-14 19:49 Last Update: 2012-06-08 19:21

Reporter:
Owner:
Type:
Status:
Closed
Component:
(None)
MileStone:
Priority:
7
Severity:
7
Resolution:
None
File:
None

Details

Mndo::RotateTwoElecTwoCoreDiatomicToSpaceFramegc. This is hotspot in SCF. Use BLAS (#27830)

Ticket History (3/6 Histories)

2012-03-14 19:49 Updated by: mikiya_fujii
  • New Ticket "Speed upMndo::RotateTwoElecTwoCoreDiatomicToSpaceFramegc" created
2012-03-14 19:49 Updated by: mikiya_fujii
  • Summary Updated
2012-03-14 19:50 Updated by: mikiya_fujii
  • Severity Update from 5 - Medium to 7
2012-06-08 18:32 Updated by: mikiya_fujii
Comment

test input is bollow, CNT+10TP

THEORY
   pm3-d
THEORY_END

SCF
   max_iter 100
   rms_density 0.00000001
   damping_thresh 10 
   damping_weight 0.98
   diis_start_error 0.01
   diis_end_error   0.0
   diis_num_error_vect 5 
SCF_END

MEMORY
   limit_heap 10000
MEMORY_END


GEOMETRY
  C 10.30477272 11.78406748 -2.27007368
  C 9.61616789 11.10092954 -1.24639799
  C 8.46089873 10.36283750 -1.58047215
  C 7.99944358 10.31012061 -2.92740881
  C 8.69054611 10.99241147 -3.94123650
  C 9.84779773 11.73083198 -3.60396786
  C 7.53986595 9.55352703 -0.73027657
  C 7.54715382 9.29022208 0.65553892
  C 6.53292505 8.48304510 1.20276465
  C 5.50547411 7.92525707 0.39171913
  C 5.51319206 8.19734253 -1.00737512
  C 6.51838482 9.00382671 -1.55641780
  C 6.71368328 9.43974670 -3.03842371
  C 5.49316531 10.28410283 -3.55083806
  C 6.93842639 8.20179640 -3.97715620
  C 4.42748305 7.06209258 1.00308402
  C 3.97680043 7.28029838 2.33309117
  C 2.98176406 6.47903915 2.91005248
  C 2.38862100 5.42165413 2.16727023
  C 2.82632771 5.19781487 0.83299158
  C 3.82715592 6.00260893 0.27057165
  C 0.38159350 5.18374707 3.67341891
  C -0.13476823 6.48436487 3.42930439
  C -1.06663668 7.05199217 4.31280821
  C -1.52480923 6.34591585 5.45094759
  C -1.01073051 5.04804578 5.68321252
  C -0.06703277 4.47160667 4.81739412
  C -2.53546581 6.98194610 6.42005850
  C 1.28331383 3.17130699 2.44815418
  C 2.46796078 2.41502057 2.23301006
  C 0.02789882 2.51043989 2.35580115
  C -0.03325331 1.14105510 2.06246405
  C 1.14320633 0.37517900 1.83952603
  C 2.39157621 1.04852612 1.93054822
  N 1.35367237 4.58982448 2.76161747
  H 10.39711410 12.26910167 -4.38817972
  H 11.20611965 12.36296198 -2.02771606
  H 9.97505492 11.14422100 -0.20926778
  H 8.34365778 10.95776724 -4.98284457
  H 8.33309555 9.70347330 1.30190154
  H 6.54379891 8.26429376 2.27821751
  H 4.71886929 7.78711566 -1.64427606
  H 4.40977923 8.09698842 2.92509469
  H 2.65166993 6.66443624 3.93973622
  H 4.15590536 5.80011348 -0.75699138
  H 2.37627925 4.38690350 0.24690968
  H 0.19895511 7.03951325 2.54363462
  H -1.45739703 8.05761067 4.10585842
  H -1.34893261 4.48161809 6.56157588
  H 0.32670559 3.46696968 5.01644172
  H -0.89446747 3.08131004 2.51917745
  H -1.01562375 0.65424729 2.01042871
  H 3.32143376 0.49214561 1.75604569
  H 3.44404879 2.91074794 2.30244903
  H -3.12769202 6.20778817 6.93129536
  H -3.22165991 7.65537907 5.88370037
  H 7.78878593 7.59807583 -3.62568120
  H 7.14574226 8.53762501 -5.00485089
  H 4.58142559 9.66709169 -3.55591459
  H 5.32497109 11.15567842 -2.90030374
  H 5.68059883 10.63993937 -4.57566779
  H 6.03806113 7.56849629 -3.99157625
  H -2.01519153 7.57577170 7.19214516
  C -0.05141209 -1.63971148 0.83204923
  C -0.13726749 -3.00965559 0.52315661
  C 0.91759177 -3.86546964 0.90547488
  C 2.04957778 -3.33983749 1.59234118
  C 2.12985264 -1.97565233 1.89698727
  C 1.07156018 -1.09852590 1.51876311
  C 1.10758272 -5.33198361 0.72414839
  C 0.29100718 -6.30705996 0.11309292
  C 0.72940419 -7.64365697 0.08260382
  C 1.97310565 -8.03436583 0.65357284
  C 2.78584557 -7.03794760 1.26918617
  C 2.35507987 -5.70609681 1.30153047
  C 3.07481080 -4.46731224 1.91122618
  C 4.45432859 -4.20235094 1.21039013
  C 3.28350677 -4.63063444 3.45862903
  C 2.42460385 -9.47478487 0.60980009
  C 2.03894100 -10.33807838 -0.45143417
  C 2.45282403 -11.67680011 -0.49549213
  C 3.28609225 -12.20699162 0.52751481
  C 3.68295254 -11.35170449 1.59192789
  C 3.25393628 -10.01780302 1.62839442
  C 4.01221557 -14.22822690 -0.79595606
  C 4.63976397 -13.49757092 -1.84097116
  C 4.90821896 -14.11470603 -3.07230793
  C 4.57896240 -15.47293976 -3.30162748
  C 3.96205000 -16.19543888 -2.25409065
  C 3.67318124 -15.58856352 -1.02026089
  C 4.87076249 -16.13199464 -4.66004096
  C 3.88222364 -14.34960785 1.72027064
  C 2.98179162 -14.15885047 2.80386418
  C 4.93748114 -15.29242467 1.85642663
  C 5.07848332 -16.02648639 3.04622040
  C 4.18812387 -15.82804847 4.12069726
  C 3.14423740 -14.88959108 3.99296969
  N 3.72474793 -13.59286957 0.48503959
  H -0.86313165 -0.96953324 0.52144434
  H -1.01322547 -3.40141676 -0.01133316
  H 2.99622475 -1.57853871 2.44122061
  H -0.67630504 -6.03251786 -0.32894647
  H 0.08825733 -8.40594250 -0.37796690
  H 3.75508768 -7.31843944 1.70075868
  H 1.40687343 -9.95225401 -1.26166899
  H 2.13855978 -12.32409837 -1.32352319
  H 3.56472367 -9.38463003 2.46935305
  H 4.32641399 -11.74572533 2.38812585
  H 4.91214917 -12.44765402 -1.67618812
  H 5.39476286 -13.53472800 -3.86822646
  H 3.69343465 -17.24924600 -2.40888477
  H 3.18626618 -16.16193926 -0.22160796
  H 5.63760470 -15.44032174 1.02488450
  H 5.89806190 -16.75197635 3.13846616
  H 4.30645813 -16.40009066 5.04977210
  H 2.44166206 -14.73238679 4.82261727
  H 2.16141276 -13.43778120 2.70177194
  H 4.79361718 -17.22694109 -4.58734561
  H 5.88207289 -15.87304056 -5.01234932
  H 2.32273059 -4.82025999 3.96069303
  H 3.72690975 -3.71556905 3.88071001
  H 5.13924587 -5.04525801 1.39019184
  H 4.32041673 -4.08284547 0.12460410
  H 4.91435117 -3.28625633 1.61189512
  H 3.96082050 -5.47403815 3.66336157
  H 4.15165609 -15.78760858 -5.42352547
  C -7.04235707 -9.82473986 0.07835719
  C -6.83233685 -8.47855089 0.44174325
  C -6.43363496 -7.56085341 -0.55343896
  C -6.24649620 -7.98732810 -1.90016276
  C -6.45851372 -9.32945021 -2.25427988
  C -6.85810171 -10.24669268 -1.25562873
  C -6.13883408 -6.10123243 -0.47244728
  C -6.16150782 -5.20756707 0.61977977
  C -5.82372149 -3.86009306 0.40743934
  C -5.45886351 -3.37298840 -0.88284450
  C -5.43233944 -4.29176049 -1.97442365
  C -5.77028900 -5.63449400 -1.76613731
  C -5.80670929 -6.79682771 -2.80160369
  C -4.39014185 -7.04612665 -3.43045787
  C -6.85544058 -6.52086305 -3.93666332
  C -5.11549349 -1.91830716 -1.08164274
  C -5.49167538 -0.94286344 -0.16093697
  C -5.16653942 0.44985963 -0.29247933
  C -4.44568045 0.98137345 -1.35969520
  C -4.02122075 0.00806926 -2.38139594
  C -4.36356679 -1.42774636 -2.24981317
  N -3.91552208 -2.22921607 -3.28058130
  N -3.30385620 0.33281784 -3.51525383
  S -3.08142119 -1.16388487 -4.37816985
  H -7.35191546 -10.55274191 0.84024779
  H -6.97639454 -8.15280980 1.48075777
  H -7.02642865 -11.29942935 -1.52031558
  H -6.31761222 -9.66955346 -3.28916574
  H -6.43098754 -5.55580365 1.62585903
  H -5.82467613 -3.17622413 1.26491397
  H -6.07194960 -1.23518389 0.72250227
  H -5.50159838 1.11511444 0.51233439
  H -5.11642047 -3.91920551 -2.95828436
  H -3.64766401 -7.24670068 -2.64335034
  H -4.06869275 -6.16091784 -4.00055705
  H -7.85345579 -6.34501407 -3.50743475
  H -6.91210130 -7.38329864 -4.61861689
  H -6.55985505 -5.63354788 -4.51737223
  H -4.42363317 -7.91001870 -4.11206150
  C -3.06295479 2.91686606 -2.28369618
  C -2.70790482 4.27629707 -2.33561548
  C -3.40936730 5.20106197 -1.53307945
  C -4.47252913 4.75497278 -0.69451527
  C -4.82274722 3.40152747 -0.64524310
  C -4.11143221 2.44937504 -1.43919318
  C -3.25738092 6.67271159 -1.36815986
  C -2.36344707 7.58860992 -1.96464846
  C -2.43707026 8.94597520 -1.60836331
  C -3.39140084 9.42062978 -0.66063447
  C -4.29974832 8.48553530 -0.08100523
  C -4.22715527 7.13190357 -0.43136282
  C -5.10378085 5.95044122 0.07860001
  C -6.61139572 6.15269794 -0.31020311
  C -4.97231981 5.76616446 1.63169361
  C -3.42641448 10.87818734 -0.26822493
  C -2.33874418 11.71994189 -0.48926705
  C -2.32450009 13.12074276 -0.13261973
  C -3.40530661 13.72897793 0.46583827
  C -4.57906391 12.92295647 0.74214214
  C -4.60010357 11.49105617 0.38494802
  N -5.77252294 10.83518352 0.70657945
  N -5.72206538 13.41517150 1.34328977
  S -6.79671433 12.04426574 1.44014310
  H -2.54712872 2.18005491 -2.91437695
  H -1.89246218 4.60781836 -2.99250480
  H -5.65191174 3.07447458 -0.00591393
  H -1.62041018 7.25103995 -2.69955079
  H -1.74930754 9.65408125 -2.08657217
  H -1.43044292 11.31110437 -0.94826384
  H -1.41673807 13.70057431 -0.34874364
  H -3.40366234 14.79167587 0.74111237
  H -5.05761563 8.86321354 0.61906999
  H -6.71700418 6.27823399 -1.39841615
  H -7.01286519 7.04849256 0.18815623
  H -3.91959362 5.62070330 1.91778736
  H -5.55169486 4.88952485 1.96013143
  H -5.35921443 6.65611637 2.15149291
  H -7.20494657 5.28017060 0.00356335
GEOMETRY_END

2012-06-08 19:12 Updated by: mikiya_fujii
  • Ticket Close date is changed to 2012-06-08 19:12
  • Status Update from Open to Closed
Comment
(This comment has been deleted)
2012-06-08 19:21 Updated by: mikiya_fujii
Comment

Completed at Rev. 727. BLAS is not used because this method had not caused cache-miss.

Benchmark is shown below (MolDS is compiled with icpc&MKL with -O2 option).
○Before speeding up (Rev. 726) in cpu time[s]
        7.889 (total: 38.287)
        8.032 (total: 38.020)
        7.752 (total: 36.470)
  -------------------------------------
  ave.  7.891 (total: 37.592)   (*1)

○After speeding up (Rev. 727) in cpu time[s]
        1.475 (total: 29.830)
        1.454 (total: 30.030)
        1.452 (total: 30.230)
  -------------------------------------
  ave.  1.460 (total: 30.030)    (*2)

○Results
  ->    5.40  (total: 1.25)      (*1)/(*2)
  ->    81.5% (total: 20.1%)     (*1 - *2)/(*1)

Attachment File List

No attachments

Edit

Please login to add comment to this ticket » Login