source: Sophya/trunk/SophyaLib/Manual/perfmachine.txt@ 3270

Last change on this file since 3270 was 3270, checked in by ansari, 18 years ago

Ajout de mesures ds perfmachine.txt (version su seminaire LAL-DevDurable) et mention de modif nt2dcn nt2dci ds modifs.tex , Reza 19/06/2007
File size: 31.9 KB
Line 
1-------------------------------------------------------------------------------------
2 Comparaison performance de differentes machines en compilation / execution (calcul)
3 -----------------------
4 Mesures effectuees en Janvier 2007 , R. Ansari / C. Magneville
5-------------------------------------------------------------------------------------
6
7(a) eros3 : Bipro-bicoeur Xeon@2.4 GHz Linux (xeon-lx-2.4GHz) , gcc 3.2
8(b) ccali : Bipro-bicoeur Xeon@2.8 GHz Linux (xeon-lx-2.8GHz) , icc 8.0 ou 9.0
9 Flags de compilation avec [-O -g]
10(c) sgsda: AMD Bipro AMD opteron 248 @ 2.2 GHz (amd-lx-
11 Flags de compilation avec [-O -g]
12(cc) grid-saclay: AMD opteron 275 Bipro-bicoeur @ 2.2 GHz (amd275-lx)
13 Flags de compilation avec [-O -g]
14
15(d) asc: bipro alpha (@ ~1 GHz) server DS20 OSF (osf1) , cxx 6.5 (osf-asc)
16(nouveau asc 420 MFLOPS, moins puissante que l'ancien asc 800 MFLOPS)
17(e) xp1000-dapnia: alpha xp1000 @ ~ 600 MHz ? OSF1 , cxx ? (osf-xp1000)
18(e') cool: alpha xp1000 @ ~ 667 MHz ? OSF1 5.1 , cxx 6.3
19(f) superosf-dapnia: multi-proc alphaServer ES80 6 procs EV7 @ 1 GHz (super-osf)
20
21(g) ccsvx01: XServe G5 bipro @~1.8-2GHz (Darwin/OSX) (G5-osx-2GHz) , gcc 3.3
22(h) PowerBook-Reza : Apple G4 @ 1.25 GHz (G4-osx-1.25GHz) , gcc 3.3
23(i) MacBook-Reza: Apple/ Core double-coeur Intel @ 1.83 GHz (core-osx-1.83GHz) gcc 4
24(j) MacPro-Grosdidier : Apple / Xeon 2 double-coeur @ 3 GHz gcc 4.0.1 , compil SOPHYA -O2 -g
25
26(p) IBM-AIX regatta , xlC , IBM eServer pSeries 655 , 8 proc power4 @ 1.1 GHz
27(q) IBM-AIX meso , AIX 5.3, xlC V8 , IBM Power5 , 8 proc bi-coeur P575 @ 1.9 GHz
28
29(s) SGI-IRIX64 magique, CC
30
31NOTES :
32- Sur les machines Xeon, il y a une interaction entre process / threads par rapport a
33l'occupation des CPU's. On perd un facteur 3 en performance multi-threads/multi-taches.
34La machine MacPro avec OSX se debrouille quand meme mieux.
35- Effet du systeme ou carte mere ???
36
37Flag de compilation
38- Flag de compilation par defaut [-O -g] en general
39- Sur eros3 (xeon-linux gcc 3.3) [-O -g] OU [-O3 -g]
40- Sur Darwin [-g] ou [-O2 -g] (ou [-tune G5] sur XServe G5)
41 Sur les mac (en particulier G4/G5), grande difference entre -g et -Ox -g
42 mais peu de difference entre -O -O2 -O3
43- Sur machine aix-meso [-O -g] ou [-O3 -g]
44
45X/ Performances brutes cpupower et donnees SPEC ((http//www.spec.org)
46----------------------------------------------------------------------
47
48(1) MFLOPS -> cpupower 2 (x/y : -O -g / -O3)
49SPECint2000 (3) / SPECfp2000 (2) (http//www.spec.org)
50
51X.1/ Performances en calcul double
52csh> cpupower 0 3000000 5
53 3 10^6 operations doubles - sur memoire 3x3 10^6 doubles (~50 MO)
54 ===> ~ 24 MO / MFLOPS
55csh> cpupower 2
56 1.6 10^9 operations doubles - sur 3x20000 doubles (~0.5 MO)
57
58
59Compilation avec -O (optimisation)
60 (1) cpupower 0 : debit memoire en MO/s
61 (2) cpupower 0 , MFLOPS
62 (5) cpupower 2 , MFLOPS
63
64Compilation avec -g (debug / sans optimisation)
65 (3) cpupower 0 , MFLOPS
66 (6) cpupower 2 , MFLOPS
67
68Compilation avec -O3 ou -fast ...( optimisation poussee)
69 (4) cpupower 0 , MFLOPS
70 (7) cpupower 2 , MFLOPS
71
72
73----------------------------------------------------------------------------------------------
74 MFLOPS |(1) MO/s| (2) (3) (4) | (5) (6) (7)
75----------------------------------------------------------------------------------------------
76(a)xeon-lx-2.4GHz | 1290 | 53 53 55 | 338 340 320
77(b)xeon-lx-2.8GHzicc | 2040 | 85 80 83 | 914 409 914
78(c)amd-lx | 1560 | 65 77 68 | 666 314 686
79(cc)amd2-lx | |
80
81(e')osf-cool | 768 | 32 15 32 | 630 150 660
82(f)superosf
83
84(g)G5-osx-1 GHz | 2100 | 88 68 88 | 1000 255 1073
85(f)G4-osx-1.25GHz | 600 | 25 16 25 | 417 93 430
86(i)core-osx-1.83GHz | 2500 | 107 75 107 | 855 309 884
87(j)xeon-osx
88
89(p)ibm-aix-regatta | 3100 | 130 55 133 | 730 115 1750 (32 bits)
90(p)ibm-aix-meso | 5700 | 240 70 320 | 1500 220 3400 (32 bits)
91
92(s)sgi-magique | 336 | 14 7 15 | 340 40 460 (32 bits)
93----------------------------------------------------------------------------------------------
94
95X.2/ Comparaison performances int, float double
96 cpupower compile avec -O
97
98(1) float , cpupowerF 0 3000000 5 / cpupowerF 2
99 -> MFLOPS (puissance de calcul sur float)
100(2) double, cpupowerD 0 3000000 5 / cpupowerF 2 (idem tableau X.1)
101 -> MDBLOPS (puissance de calcul sur float)
102(3) int, cpupowerI 0 3000000 5 / cpupowerI 2
103 -> MINTOPS (puissance de calcul sur int=4 bytes)
104(4) long (ou long long (*)) cpupowerL 0 3000000 5 / cpupowerL 2
105 -> MLONOPS (puissance de calcul sur long=8 bytes)
106----------------------------------------------------------------------------------------------
107 MFLOPS | (1)MFLOPS (2)MDBLOPS (3)MINTOPS (4)MLONOPS
108----------------------------------------------------------------------------------------------
109(a)xeon-lx-2.4GHz |
110(a)xeon-lx-2.8GHzicc | 166/905 90/900 166/1500 88/522 (*)
111(c)amd-lx | 125/695 65/675 125/1570 65/1045
112(cc)amd2-lx |
113
114(e')osf-cool | 60/635 32/631 62/640 31/630
115(f)superosf
116
117(g)G5-osx-1 GHz | 180/1260 90/1150 165/940 81/280 (*)
118(f)G4-osx-1.25GHz | 45/430 25/410 45/710 24/190 (*)
119(i)core-osx-1.83GHz | 185/919 105/855 187/935 62/246 (*)
120(j)xeon-osx
121
122(p)ibm-aix-regatta |
123(p)ibm-aix-meso | 250/1150 250/1500 250/1200 50/200 (32 bits)
124 | 280/1500 250/1600 250/1100 210/1000 (64 bits -q64)
125
126(s)sgi-magique |
127----------------------------------------------------------------------------------------------
128
129X.3/ Comparaison avec SPEC
130csh> cpupower 0 / cpupower 2
131----------------------------------------------------------------------
132 MFLOPS(1) SPECfp SPECint
133----------------------------------------------------------------------
134(b)xeon-lx-2.8GHz 166/900 1400 1400
135(c)amd-lx 125/690 1600 1300
136(cc)amd2-lx 675 1600 1300
137
138(e)osf-xp1000 32/650 500 400
139(f)superosf 842 1100 700
140
141(i)core-osx-1.83GHz 110/850 1400 1500
142(j)xeon-osx 2600 2900 -
143
144(p)ibm-aix-regatta 130/700 1050 700
145----------------------------------------------------------------------
146
147
148A/ Compilation tout SOPHYA
149----------------------------
150csh> time make all (1)
151ou
152csh> time make -j 2 all (2)
153 Temps CPU
154 Indice de performance 100*(1000/TCPU)
155 Temps elapsed (vrai)
156 Temps vrai / TCPU
157
158
159----------------------------------------------------------------------
160 CPU(s) IndPerf TElapsed , TCPU/Elapsed %
161----------------------------------------------------------------------
162(a)xeon-lx-2.4GHz (2) 615 s 162 410 s 150%
163 avec -O3 -g (2) 1300 s 77 760 s 172%
164(b)xeon-lx-2.8GHz (2) 755 s 132 540 s 140%
165(c)amd-lx (2) 336 s 297 175 s 192%
166
167(d)osf-asc (1) 1920 s 52 2340 s 83% (??)
168(e)osf-xp1000 (1) 533 s 187 660 s 80%
169(f)superosf (1) 895 s 112 910 s 98%
170
171(g)G5-osx-2GHz (2) 453 s 221 250 s 182%
172 -tune=G5 1100 s 90
173 -g -O 740 s 380 s 195%
174(h)G4-osx-1.25GHz (1) 660 s 151 710 s 93% [-g]
175 1500 s 94% [-O2 -g]
176(i)core-osx-1.83GHz (2) 209 s 478 116 s 180%
177 -O2 (1) 367 s 272 381 96%
178(j)xeon-osx
179
180(p)ibm-aix
181----------------------------------------------------------------------
182
183Taille shared libs :
184(a)
185(c) 33 MO
186(f) = (e) = 57 MO
187(g) 80 MO
188(i) 83 MO
189
190B/ Calcul brut (Tableaux de SOPHYA) avec / sans threads
191--------------------------------------------------------
192B.1.a/ Calcul sur vecteur 10 * V2 ~= DLO4 (V1)
193 ~ 10 x 10 x 9. 10^6 operations double sur 2 x 9 10^6 double
194 900 M.Ops r_8 / ~ 1500 MO
195
196(1) time cpupower 0 # compile avec -O (/ -O -g)
197(2) time zthr arrdl 1 3000 1 thread
198(3) time zthr arrdl 2 3000 2 thread
199(4) time zthr arrdl 4 3000 4 thread
200(5) time zthr arrdl 6 3000 6 thread
201(6) time zthr arrdl 8 3000 8 thread
202
203-----------------------------------------------------------------------------------
204 (1)MFLOPS (2)CPU/Elap/% (3)CPU/Elap/% (4)CPU/Elap/%
205-----------------------------------------------------------------------------------
206(a)xeon-lx-2.4GHz 53
207(b)xeon-lx-2.8GHz 65 2.6/2.6/100% 5.3/2.9/180% 14.3/4.86/310%
208 (5) 23/7.4/314%
209(c)amd-lx 95
210
211
212(e')osf-cool 32 5.7/5.8/98% 11.1/11.3/98% 22.3/22.5/98%
213(f)superosf
214
215(g)G5-osx-2GHz 88 2.5/2.6/99% 5.9/3.38/184% 11/6.45/173% [-O2 -g]
216(h)G4-osx-1.25GHz 25 6.6/7/95% 13.4/13.8/97% [-O2 -g]
217(i)core-osx-1.83GHz 107 2.1/2.1/98% 4.3/2.9/150% 8.3/30/31% [-O2 -g]
218(j)xeon-osx
219
220(p)ibm-aix-regatta 130
221(q)ibm-aix-meso 150 0.7/1/70% 1.2/2./60% 3.2/2/150% [-O3]
222 (5) 5.4/3/180% (6) 6.4/3/210%
223
224(s)sgi-magique 7 78/78/99% 167/95/175% 339/96/352% [-O -g: NON-OPT]
225 14 16.4/16.5/99% 33.8/22.4/150% 79/32/250% [-O -g2 OPT]
226 -----------------------------------------------------------------------------------
227
228B.1.b/ Calcul sur vecteur V2 = Sin(V1) + Cos(V1)
229 ~ 50 x 9. 10^6 operations double sur 2 x 9 10^6 double, mem ~ 150 MO
230 ~500 M.Ops r_8 / ~ 600 MO I/O
231
232(1) time cpupower 0 # compile avec -O (/ -O -g)
233(2) time zthr arrmf 1 3000 1 thread
234(3) time zthr arrmf 2 3000 2 thread
235(4) time zthr arrmf 4 3000 4 thread
236(5) time zthr arrmf 6 3000 6 thread
237(6) time zthr arrmf 8 3000 8 thread
238
239-----------------------------------------------------------------------------------
240 (1)MFLOPS (2)CPU/Elap/% (3)CPU/Elap/% (4)CPU/Elap/%
241-----------------------------------------------------------------------------------
242(a)xeon-lx-2.4GHz 53
243(b)xeon-lx-2.8GHz 65 1.7/1.7/100% 3.5/2.1/173% 9.8/3.6/275%
244(c)amd-lx 95
245
246(e')osf-cool 32 4.2/4.3/98% 8.2/8.4/98% 16.1/16.2/98%
247(f)superosf
248
249(g)G5-osx-2GHz 88 2.3/2.3/100% 5/3/165% 9.6/5.8/167% [-O2 -g]
250(h)G4-osx-1.25GHz 25 4.5/4.8/95% 10.9/14.6/72% [-O2 -g]
251(i)core-osx-1.83GHz 107 2.3/2.3/98% 4.8/3.1/158% [-O2 -g]
252(j)xeon-osx
253
254(p)ibm-aix-regatta 130
255(q)ibm-aix-meso 150 1./2/50% 2.8/3/86% 5.4/4/130% [-O3]
256 (5) 10/4/250% (6) 11.2/5/220%%
257
258(s)sgi-magique 7 11.5/11.7/99% 24/17/140% 51.5/18.4/280% [-O -g NON-OPT]
259 14 6.5/6.6/99% 13.3/12/110% 34.5/17.3/200% [-O -g3 OPT]
260 -----------------------------------------------------------------------------------
261
262
263B.1.c/ Version corrige de zthr.cc (apres 23/05/07)
264 arr = (c1*a1) + (c2*a2)
265 ~ 3 x 4. 10^6 operations int_4 sur 3 x 4 10^6 int_4
266 12 M.Ops int_4 / ~ 50 MO
267
268(1) time cpupower 0 # compile avec -O (/ -O -g)
269(2) time zthr arr 1 2000 1 thread
270(3) time zthr arr 2 2000 2 thread
271(4) time zthr arr 4 2000 4 thread
272(5) time zthr arr 6 2000 6 thread
273(6) time zthr arr 8 2000 8 thread
274
275-----------------------------------------------------------------------------------
276 (1)MFLOPS (2)CPU/Elap/% (3)CPU/Elap/% (4)CPU/Elap/%
277-----------------------------------------------------------------------------------
278(a)xeon-lx-2.4GHz 53 0.5/1/43% 1/1.1/88% 2.8/1/262%
279 (5) 4.5/1.8/246% (6) 6.1/2.1/310%
280
281
282(b)xeon-lx-2.8GHz 65
283
284(c)amd-lx 95 0.23/1/22% 0.44/1/51% 1/1/102% [-O -g]
285 (5) 1.6/1/106% (6) 2.2/1.2/100%
286
287
288
289(e')osf-cool 32 0.43/1.2/35% 0.6/1.33/44% 1.1/1.3/82% [-O -g]
290 (5) 1.45/1.7/85% (6) 1.83/2.16/84%
291(f)superosf
292
293(g)G5-osx-2GHz 88 1.5/1.5/100% 3.2/1.7/185% 6.6/3.5/188% [-O -g]
294(g)G5-osx-2GHz 88 0.4/1/40% 0.9/1.0/90% 2/1.2/169% [-tune=G5 -g]
295 (5) 3.3/2/164% (6) 4.3/2.6/165%
296(h)G4-osx-1.25GHz 25 3/3/95% [-O2 -g]
297
298(i)core-osx-1.83GHz [-O2 -g]
299
300(j)xeon-osx
301
302
303(p)ibm-aix-regatta 130
304
305(q)ibm-aix-meso 150 0.6/1/58% 1/1/91% 1.7/1.2/132% [-O3]
306 (5) 2.4/1.2/193% (6) 4.25/1.6/265%
307
308 -----------------------------------------------------------------------------------
309
310B.1.x/ ancienne version de zthr (avant 23/05/07)
311 On faisait 2 multiplications par ctye suivi d'un produit matriciel !
312 arr = c1*a1*c2*a2 ( ~ 3 10^6 op. double)
313(1) time cpupower 2 # compile avec -O3 (/ -O -g)
314(2) time zthr arr 1 1000 1 thread
315(3) time zthr arr 2 1000 2 thread
316(4) time zthr arr 4 1000 4 thread
317(5) time zthr arr 6 1000 6 thread
318(6) time zthr arr 8 1000 8 thread
319
320-----------------------------------------------------------------------------------
321 (1)MFLOPS (2)CPU/Elap/% (2)IndPerf (3)CPU/Elap/% (3)IndPerf
322-----------------------------------------------------------------------------------
323(a)xeon-lx-2.4GHz 1167 5.15/5.2/99% 11.4/5.8/196%
324 (4)36.6/9.28/394%
325 -O3 -g 4.9/5./99%
326(b)xeon-lx-2.8GHz 920 2.3/2.3/100% 6.2/3.1/198%
327 (4)26/6.6/396%
328(c)amd-lx 690 3.6/3.6/99% 6.8/4/171%
329 (4)13.5/7/193%
330 (5)20.3/10.23/198%
331(cc)amd2-lx 675 2/2/99% 4.15/2.1/197%
332 (4)8.25/4.15/198%
333 (5)13.6/4.6/292%
334 (6)19.8/6.5/300%
335
336(d)osf-asc 420 6.3s/6.5s/99% 16.9/8.8/192%
337 (4)29.9/15.7/191%
338(e)osf-xp1000 648 5.1/5.3/96.6% 11.4/11.4/99%
339 (4)25.2/25.5/99%
340(f)superosf 842 2.87/2.88/99.6% 6.25/4.1/153%
341 (4)11.6/3.06/379%
342
343(h)G4-osx-1.25GHz 92 44s/48s/91% 86.7/99.8/92% [-g]
344 380 12.2/12.9/95% 24/25.3/95% [-O2 -g]
345(g)G5-osx-2GHz 1151 20s/20s/99% 40s/23s/170%
346 (4) 80.8/45/180%
347 -O -g 4.5/4.9/91% 9.3/4.7/197%
348 (4) 18.3/9.4/197%
349 -tune=G5 3.35/3.8/88% 7.1/3.6/196%
350(h)G4-osx-1.25GHz 92 44s/48s/91% 86.7/99.8/92% [-g]
351 380 12.2/12.9/95% 24/25.3/95% [-O2 -g]
352 (4) 14/7.5/187%
353(i)core-osx-1.83GHz 855 11.5/11.5/100% 23/11.6/192% [-g]
354 (4) 46/23/199%
355 -O2 3.85/3.89/99% 7.7/3.9/198% [-O2 -g]
356 (4) 15.4/7.77/198%
357
358(j)xeon-osx 2600 2.5/2.5/100% 5.1/2.6/199%
359 (4) 11.5/3.2/362%
360 (5) 17.4/4.77/365%
361
362(p)ibm-aix-regatta 1750/730 6.8/6.9/98% 13.1/6.75/195%
363 (4) 26.3/11.7/225%
364(q)ibm-aix-meso 3600/1250 3.6/3.75/96% 7.35/3.7/197%
365 (4) 12.46/4.2/298%
366 (5) 219/6.7/280%
367 (6) 24/4.5/530%
368
369
370(s)sgi-magique 460 60/60/99%
371 -----------------------------------------------------------------------------------
372
373
374B.2/ Multiplication de matrices mtx = mtx1 * mtx2
375 ~ 2 10^9 op. double / thread
376(1) time cpupower 2 (-O3 / -O -g)
377(2) time zthr mtx 1 1000 1 thread
378(3) time zthr mtx 2 1000 2 thread
379(4) time zthr mtx 4 1000 4 thread
380(5) time zthr mtx 6 1000 6 thread
381(6) time zthr mtx 8 1000 8 thread
382
383-----------------------------------------------------------------------------------
384 (1)MFLOPS (2)CPU/Elap/% (2)IndPerf (3)CPU/Elap/% (3)IndPerf
385-----------------------------------------------------------------------------------
386(a)xeon-lx-2.4GHz 1167 6.5/6.5/100% 17.4/8.8/198%
387 (4) 80.5/20.3/397%
388 (5) 114.5/29.6/387%
389 (6) 160/40.3/387%
390
391(b)xeon-lx-2.8GHz 920 3.4/3.4/100% 12/6.1/199%
392 (4) 55.8/14/400%
393 (5) 79.5/20.3/392%
394 (6) 102/25.8/396%
395(c)amd-lx 690 6.98/6.98/100% 14.1/8.15/173%
396 (4) 27.7/14.23/194%
397 (5) 41.4/21.07/196%
398 (6) 55.4/27.9/198.7%
399(cc)amd2-lx 675 4.1/4.1/100% 9.55/4.8/198%
400 (4) 20/10.27/195%
401 (5) 32.8/11.16/294%
402 (6) 42.75/13.8/309%
403
404
405(d)osf-asc 420 13.5s/13.7s/98% 32/16.5/194%
406 (4) 67.5/34.4/196%
407(e)osf-xp1000 648 13/14.1/92% 27.1/27.4/99%
408 (4) 54/54.7/99.6%
409 (5) 80.6/81/99.6%
410 (6) 107.8/108.3/99.5%
411(e')osf-cool 13/13.22/98% 26/26.1/99%
412 (4) 51.8/51.9/99%
413(f)superosf 842 6.1/7.24/84% 12.35/6.29/196%
414 (4) 24.3/6.31/385%
415 (5) 36.5/10.9/335%
416 (6) 50.1/18.15/276%
417
418(g)G5-osx-2GHz 1151 23/23.7/97% 46.5/27.5/170%
419 (4) 93.4/49.4/189%
420 -O -g 6.2/6.2/100% 14.2/7.2/197%
421 (4) 28.3/14.36/197%
422 -tune=G5 5.7/5.8/98% 13.3/6.8/197%
423 (4) 26.8/13.56/197%
424 (6) 53.8/27.25/197%
425(h)G4-osx-1.25GHz 333 23.5/24.5/96% [-O2]
426(i)core-osx-1.83GHz 855 12.6/12.7/100% 25.8/13.4/194%
427 (4) 51.6/26/199%
428 -O2 4.25/4.5/94% 10.6/5.36/198%
429 (4) 20.87/10.68/198%
430 -O2 2 jobs // 2 x 5/5.4/92%
431(j)xeon-osx 2600 2.8/2.8/99% 9.3/4.66/199%
432 (4) 31.4/8.6/364%
433 (5) 47.1/12.96/364%
434 (6) 62.8/17.38/362%
435
436(p)ibm-aix-regatta 1750/730 9.5/9.7/98% 18.3/16.0/114%
437 (4) 38.3/24.7/155%
438(p)ibm-aix-meso 3600/1250 2.3/2.3/99% 5.1/2.64/194% (compil avec -O3)
439 (4) 11.4/4.16/272%
440 (5) 20.2/5.85/344%
441 (6) 29.9/6.74/442%
442
443(s)sgi-magique 400 44/44.3/99% 96.5/55/176%
444
445 -----------------------------------------------------------------------------------
446
447
448B.4/ Operations sur tableaux doubles- mesures avec spar
449 csh> time spar 2 1 2000 2000
450 (1) cpupower 2 MFLOPS
451 (2) MFLOPS (double) spar
452 (3) time spar 2 5 1000 2000 CPU/Elap/%
453-----------------------------------------------------------------------------------
454 (1)MFLOPS (2)CPU / % (3)CPU/Elap/%
455-----------------------------------------------------------------------------------
456(a)xeon-lx-2.4GHz 53 ~ 20-35 MFLOPS , 90% 20/20.2/99% [-g -O]
457
458
459(b)xeon-lx-2.8GHz 65
460
461(c)amd-lx 95 ~ 20-40 MFLOPS , 99% 17.2/17.2/100% [-g -O]
462
463
464(d)osf-asc
465
466(e)osf-xp1000 32 ~ 15-25 MFLOPS , 90% 37.6/41.2/91% [-g -O]
467(f)superosf
468
469(g)G5-osx-2GHz 88 ~ 10-25 MFLOPS , 99% 45/45/100% [-g -O] ou [-g -O2]
470(h)G4-osx-1.25GHz 25 ~ 8-16 MFLOPS , 92% 45.5/52/90% [-g -O2]
471
472(i)core-osx-1.83GHz
473
474(j)xeon-osx
475
476
477(p)ibm-aix-regatta 130
478
479(q)ibm-aix-meso 150 ~ 80-100 MFLOPS , 90% 5./23/22% [-O3]
480
481
482
483(s)sgi-magique 460
484 -----------------------------------------------------------------------------------
485
486B.5/ Calcul/comparaison avec JET/tjet
487csh> time tjet 10 2000 2000 OU tjet 10 2000 1000
488 (1) TCPU EltAccess C/pointeurs
489 (2) TCPU m1*c1+m2*c2+m3*c3 C/pointeurs
490 (3) TCPU EltAccess SOPHYA
491 (4) TCPU m1*c1+m2*c2+m3*c3 SOPHYA / Methodes (MultCst, AddArr ...)
492 (5) TCPU m1*c1+m2*c2+m3*c3 SOPHYA JET
493
494 -----------------------------------------------------------------------------------
495 (1) (2) (3) (4) (5)
496 -----------------------------------------------------------------------------------
497(b)xeon-lx-2.8GHz-icc 0.87 0.63 1.55 2.7/1.6 0.57
498(c)amd-lx 0.94 0.79 1.85 3.4/2.1 0.76
499
500(e')osf-cool 2.85 2.45 3.1 6.5/5.5 4.1
501
502(g)G5-osx-2GHz 1.5 0.61 2.1 4.1/1.6 0.6 (-g -O2)
503 -tune=G5 -fast 1.1 0.62 1.3 4.1/1.6 0.58
504(h)G4-osx-1.25GHz 3.86 2.2 5 9.4/6.2 3 (-g -O2)
505(i)core-osx-1.83GHz 1.1 0.49 1.6 2.8/1.7 0.68
506
507(q)ibm-aix-meso 0.43 0.27 0.52 1.12/0.75 0.35
508
509(s)sgi-magique 2.45 1.9 5.65 7.45/6.3 2.8 (-O -g3)
510-----------------------------------------------------------------------------------
511
512
513C/ Calcul fft (FFTW , FFTPack )
514-------------------------------
515
516(1) time cpupower 2
517(2) time tfft 2000000 W d 0 0 (avec FFTW)
518(3) time tfft 2000000 P d 0 0 (avec FFTPack_Sophya)
519
520IndPerf=1000/TCPU
521
522
523-----------------------------------------------------------------------------------
524 (1)MFLOPS (2)CPU/Elap/% (2)IndPerf (3)CPU/Elap/% (3)IndPerf
525-----------------------------------------------------------------------------------
526(a)xeon-lx-2.4GHz 1167 5.5/5.6/97% 180 7.4/7.4/100% 135
527 -O3 -g 6.8/7.1/96% 147
528
529(b)xeon-lx-2.8GHz 920 3.6/3.7/98% 3.7/3.8/99%
530 ~2x 6.5/8.8/73%
531 ~4x 8/14/55% (15 sec elapsed)
532
533(c)amd-lx 690 3.2/3.75/86% 4.2/4.2/100% 238
534 ~2x 4.7/4.7/99%
535(cc)amd2-lx 675 2.8/2.8/99% 3.56/3.58/99%
536
537(d)osf-asc 420 13.3/13.9/95.7% 75 12.2/17.6/70% 82
538(e)osf-xp1000 648 9.9/10.2/97% 101 9.3/9.46/98.5% 107
539(e')cool 9.2/10.1/91% 9.1/9.3/98%
540(f)superosf 842 6./6.22/96.7% 166 5.1/5.18/97.4% 196
541
542(g)G5-osx-2GHz 1151 8.5/8.7/97.5% 120 11.5/11.6/99% 87
543 -O -g 5.1/5.2/98% 190 5.4/5.5/99% 180
544 -tune=G5 5/5.1/98% 200 5.5/5.6/97% 180
545(h)G4-osx-1.25GHz 92 15.2/15.9/96% 66 23.8/34.1/70% 42 [-g]
546 380 8.8/10.3/86% 14.7/20.9/65% [-O2 -g]
547(i)core-osx-1.83GHz 855 4.6/4.75/97% 217 7/7.08/99% 142
548 -O2 3.2/3.2/99% 312 3.6/3.6/99% 277
549 -O2 2 jobs // 2 x 3.9/4.3/90% 2x 4.7/5/91% 200
550(j)xeon-osx 2600 2.6/2.6/98% 384
551 2 jobs // ~2x 3.6/4.1/87% 250
552 4 jobs //
553
554(p)ibm-aix-regatta 1750/730 6.25/18.9/33% 160 5.25/15.7/33% 190
555(q)ibm-aix-meso 3600/1250 3.95/4.3/91% 250 3.82/4./94% 260
556 2 jobs // ~2x 3.88/4.2/92% 250
557 2.8/4.76/59% 2.1/4.45/47% [-O3]
558 ~2x 2.5/4.7/54%
559 ~4x 2.5/4.8/55%
560
561
562(s)sgi-magique 460 22/22/98% 42 24.5/25/99% 40
563 -----------------------------------------------------------------------------------
564
565
566D/ Calcul inversion par lapack
567-------------------------------
568
569lpk inverse 1000,1000 0
570---> temps de calcul inversion par lapack
571-------------------------------------------------------------------------------------------
572 CPU/Elap/% (1)
573-------------------------------------------------------------------------------------------
574(a)xeon-lx-2.4GHz 5.6/~100%
575(b)xeon-lx-2.8GHz 5.34/~90%
576(c)amd-lx 5.5/5.5/99%
577
578(d)osf-asc
579(e')cool 2.8/2.9/95%
580(f)superosf
581
582(f)G4-osx-1.25GHz 2.3/~100% [-O2 -g]
583(h)G5-osx-2GHz 0.8/~100%
584 -O -g 0.86/~100%
585(i)core-osx-1.83GHz 1.93/~100%
586 -O2
587(p)ibm-aix-regatta
588(q)ibm-aix-meso 0.55/~100%
589
590(s)sgi-magique 5.3/~90%
591--------------------------------------------------------------------------------------------
592
593
594K/ Efficacite de gestion de lock (mutex) avec les threads et tableaux
595----------------------------------------------------------------------
596(32 threads - operant sur 2000 vecteurs ~ 64000 lock/unlock/wait/broadcast)
597
598(1) time zthr syncp 32 2000 4
599(2) time zthr sync 32 2000 4
600(1) time zthr syncp 4 15000 130
601(2) time zthr sync 4 15000 130
602-------------------------------------------------------------------------------------------
603 CPU/Elap/% (1) (2) (3) (4)
604-------------------------------------------------------------------------------------------
605(a)xeon-lx-2.4GHz 23.5/14/168% 4.3/1.2/365% 7.9/5.5/142% 4/2.15/190%
606 Avant ThSafeOp 17/178%
607 avec -O3 -g
608(b)xeon-lx-2.8GHz (2)
609(c)amd-lx 0.6/1/63% 0.6/1/60% 3.5/3.5/102% 2.6/2.7/98%
610
611(d)osf-asc 4.5/3.4/132% 3.35/2/170% 15.8/10.5/150% 13/8/163%
612 5.4/100%(NoThSafe)
613(e')cool 1.3/1.37/95% 1.35/1.5/89% 5.3/5.3/99% 5.2/5.2/99%
614(e)superosf (1)
615
616
617(g)G5-osx-2GHz (2) 2.6/130% (NoThSafe)
618 -O -g 2.6/1.7/150% 6.8/3.7/187% 4/2.75/142% 4.7/3/155%
619(h)G4-osx-1.25GHz (1) 40.5/42.6/95% 42.2/43.5/97% [-g]
620 3.9/4.3/89% 3.5/4/89% 3.8/4/95% 4.3/4.6/93%
621
622(i)core-osx-1.83GHz 7.7/7.1/108% 7.8/6.7/116% 30.2/29.6/102% 30.3/29.2/104% [-g]
623 -O2 2.7/1.8/152% 6/3.16/190% 3.4/2.4/142% 3.2/2.5/164% [-O2 -g]
624(j)xeon-osx
625 Avant ThSafeOp 2.55/143%
626
627(p)ibm-aix-regatta 4.7/111%
628(q)ibm-aix-meso 7.5/2.8/300% 17/3.8/450% 8.2/3.05/270% 4.85/2.43/200%
629
630--------------------------------------------------------------------------------------------
631
632
633
634L/ I/O et PPF
635-----------------
636Ecriture/lecture de n=10^7 lignes de int+6double, Total ~ 500 MO
637(1) time tstdtable w xx.ppf swap 10000000 1024 0
638(2) time tstdtable r xx.ppf swap 10000000 1024 0
639(3) time tstdtable w xx.ppf swap 50000000 1024 0
640(4) time tstdtable r xx.ppf swap 50000000 1024 0
641
642-------------------------------------------------------------------------------------------
643 CPU/Elap/% (1) (2) (3) (4)
644-------------------------------------------------------------------------------------------
645(a)xeon-lx-2.4GHz 17/26/63% 5.5/5.6/94%
646
647(b)xeon-lx-2.8GHz 7/18.5/40% 2.7/2.8/100% 7000000
648
649(c)amd-lx 5.9/6./97% 3.4/3.4/100% 30/32/93% 24/165/13% ?
650
651(d)osf-asc
652(e')cool 15/30/50% 13/13/99%
653(f)superosf
654
655(g)G5-osx-2GHz 14/14.2/98% 6/6.14/99%
656
657(h)G4-osx-1.25GHz 26/29.7/87% 15.7/38.6/41% [-O2 -g]
658(i)core-osx-1.83GHz 37/37.8/98% 22/48.4/45% [-g]
659 -O2 10.5/17.4/55% 11.2/41.2/27% [-O2 -g]
660(p)ibm-aix-regatta
661(q)ibm-aix-meso 5.5/16.8/38% 5.7/13.2/43% 32.7/85/39% 29/60/49%
662 6/11/55% 4/9/44%
663 6.3/10.6/60% 4/6/64%
664 2 lecture // ~2x 4/6/60% Elapsed 7 sec
665 1 write + 2 read // 6.3/10.4/60% ~2 3.8/6.4/59% Elapsed < 11 sec
666 1 write + 3 read // 19 sec
667 4 read // 6 sec
668--------------------------------------------------------------------------------------------
669
670
671
Note: See TracBrowser for help on using the repository browser.