Context Navigation

source: Sophya/trunk/SophyaLib/Manual/perfmachine.txt@ 3159

Visit:

Last change on this file since 3159 was 3159, checked in by ansari, 19 years ago
Modifs fichiers comparaison perfo machines, (compil avec -O2 sur certaines, + tests FFT avec 2 jobs en ), Reza 29/01/2007
File size: 10.1 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	(c) sgsda: AMD (amd-lx-
10
11	(d) asc: bipro alpha (@ ~1 GHz) server DS20 OSF (osf1) , cxx 6.5 (osf-asc)
12	(nouveau asc 420 MFLOPS, moins puissante que l'ancien asc 800 MFLOPS)
13	(e) xp1000-dapnia: alpha xp1000 @ ~ 600 MHz ? OSF1 , cxx ? (osf-xp1000)
14	(f) superosf-dapnia: multi-proc alphaServer ES80 6 procs EV7 @ 1 GHz (super-osf)
15
16	(g) ccsvx01: XServe G5 bipro @~1GHz (Darwin/OSX) (G5-osx-1GHz) , gcc 3.3
17	(h) PowerBook-Reza : Apple G4 @ 1.25 GHz (G4-osx-1.25GHz) , gcc 3.3
18	(i) MacBook-Reza: Apple/ Core double-coeur Intel @ 1.83 GHz (core-osx-1.83GHz) gcc 4
19	(j) MacPro-Grosdidier : Apple / Xeon 2 double-coeur @ 3 GHz gcc 4.0.1 , compil SOPHYA -O2 -g
20
21	(p) IBM-AIX regatta , xlC ,
22
23
24	NOTES :
25	- Sur les machines Xeon, il y a une interaction entre process / threads par rapport a
26	l'occupation des CPU's. On perd un facteur 3 en performance multi-threads/multi-taches.
27	La machine MacPro avec OSX se debrouille quand meme mieux.
28	- Effet du systeme ou carte mere ???
29
30	A/ Compilation tout SOPHYA
31	----------------------------
32	csh> time make all (1)
33	ou
34	csh> time make -j 2 all (2)
35	Temps CPU
36	Indice de performance 100*(1000/TCPU)
37	Temps elapsed (vrai)
38	Temps vrai / TCPU
39
40
41	----------------------------------------------------------------------
42	CPU(s) IndPerf TElapsed , TCPU/Elapsed %
43	----------------------------------------------------------------------
44	(a)xeon-lx-2.4GHz (2) 615 s 162 410 s 150%
45	avec -O3 -g (2) 1300 s 77 760 s 172%
46	(b)xeon-lx-2.8GHz (2) 755 s 132 540 s 140%
47	(c)amd-lx
48
49	(d)osf-xp1000 (1) 533 s 187 660 s 80%
50	(e)superosf (1) 895 s 112 910 s 98%
51	(f)osf-asc (1) 1920 s 52 2340 s 83% (??)
52
53	(f)G4-osx-1.25GHz (1) 660 s 151 710 s 93%
54	(h)G5-osx-1GHz (2) 453 s 221 250 s 182%
55	-tune=G5 1100 s 90
56	(i)core-osx-1.83GHz (2) 209 s 478 116 s 180%
57	-O2 (1) 367 s 272 381 96%
58	(j)xeon-osx
59
60	(p)ibm-aix
61	----------------------------------------------------------------------
62
63	Taille shared libs :
64	(a)
65	(f) = (e) = 57 MO
66	(h) 80 MO
67	(i) 83 MO
68
69	B/ Calcul brut avec / sans threads
70	-----------------------------------
71
72	B.1/ arr = c1a1+c2a2
73	(1) time cpupower 2 # compile avec -O3
74	(2) time zthr arr 1 1000 1 thread
75	(3) time zthr arr 2 1000 2 thread
76	(4) time zthr arr 4 1000 4 thread
77	(5) time zthr arr 6 1000 6 thread
78	(6) time zthr arr 8 1000 8 thread
79
80	-----------------------------------------------------------------------------------
81	(1)MFLOPS (2)CPU/Elap/% (2)IndPerf (3)CPU/Elap/% (3)IndPerf
82	-----------------------------------------------------------------------------------
83	(a)xeon-lx-2.4GHz 1167 5.15/5.2/99% 11.4/5.8/196%
84	(4)36.6/9.28/394%
85	-O3 -g 4.9/5./99%
86	(b)xeon-lx-2.8GHz 920 2.3/2.3/100% 6.2/3.1/198%
87	(4)26/6.6/396%
88	(c)amd-lx 690 3.6/3.6/99% 6.8/4/171%
89	(4)13.5/7/193%
90	(5)20.3/10.23/198%
91
92	(d)osf-xp1000 648 5.1/5.3/96.6% 11.4/11.4/99%
93	(4)25.2/25.5/99%
94	(e)superosf 842 2.87/2.88/99.6% 6.25/4.1/153%
95	(4)11.6/3.06/379%
96	(f)osf-asc 420 6.3s/6.5s/99% 16.9/8.8/192%
97	(4)29.9/15.7/191%
98
99	(f)G4-osx-1.25GHz 333 44s/48s/91% 86.7/99.8/92%
100	(h)G5-osx-1GHz 1151 20s/20s/99% 40s/23s/170%
101	(4) 80.8/45/180%
102	-tune=G5 3.35/3.8/88% 7.1/3.6/196%
103	(4) 14/7.5/187%
104	(i)core-osx-1.83GHz 855 11.5/11.5/100% 23/11.6/192%
105	(4) 46/23/199%
106	-O2 3.85/3.89/99% 7.7/3.9/198%
107	(4) 15.4/7.77/198%
108
109	(j)xeon-osx 2600 2.5/2.5/100% 5.1/2.6/199%
110	(4) 11.5/3.2/362%
111	(5) 17.4/4.77/365%
112
113	(p)ibm-aix-regatta 700 6.8/6.9/98% 13.1/6.75/195%
114	(4) 26.3/11.7/225%
115
116	-----------------------------------------------------------------------------------
117
118
119	B.2/ Multiplication de matrices mtx = mtx1 * mtx2
120
121	(1) time cpupower 2
122	(2) time zthr mtx 1 1000 1 thread
123	(3) time zthr mtx 2 1000 2 thread
124	(4) time zthr mtx 4 1000 4 thread
125	(5) time zthr mtx 6 1000 6 thread
126	(6) time zthr mtx 8 1000 8 thread
127
128	-----------------------------------------------------------------------------------
129	(1)MFLOPS (2)CPU/Elap/% (2)IndPerf (3)CPU/Elap/% (3)IndPerf
130	-----------------------------------------------------------------------------------
131	(a)xeon-lx-2.4GHz 1167 6.5/6.5/100% 17.4/8.8/198%
132	(4) 80.5/20.3/397%
133	(5) 114.5/29.6/387%
134	(6) 160/40.3/387%
135
136	(b)xeon-lx-2.8GHz 920 3.4/3.4/100% 12/6.1/199%
137	(4) 55.8/14/400%
138	(c)amd-lx 690 6.98/6.98/100% 14.1/8.15/173%
139	(4) 27.7/14.23/194%
140	(5) 41.4/21.07/196%
141	(6) 55.4/27.9/198.7%
142
143	(d)osf-xp1000 648 13/14.1/92% 27.1/27.4/99%
144	(4) 54/54.7/99.6%
145	(5) 80.6/81/99.6%
146	(6) 107.8/108.3/99.5%
147	(e)superosf 842 6.1/7.24/84% 12.35/6.29/196%
148	(4) 24.3/6.31/385%
149	(5) 36.5/10.9/335%
150	(6) 50.1/18.15/276%
151	(f)osf-asc 420 13.5s/13.7s/98% 32/16.5/194%
152	(4) 67.5/34.4/196%
153
154	(f)G4-osx-1.25GHz 333
155	(h)G5-osx-1GHz 1151 23/23.7/97% 46.5/27.5/170%
156	(4) 93.4/49.4/189%
157	-tune=G5 5.7/5.8/98% 13.3/6.8/197%
158	(4) 26.8/13.56/197%
159	(6) 53.8/27.25/197%
160	(i)core-osx-1.83GHz 855 12.6/12.7/100% 25.8/13.4/194%
161	(4) 51.6/26/199%
162	-O2 4.25/4.5/94% 10.6/5.36/198%
163	(4) 20.87/10.68/198%
164	-O2 2 jobs // 2 x 5/5.4/92%
165	(j)xeon-osx 2600 2.8/2.8/99% 9.3/4.66/199%
166	(4) 31.4/8.6/364%
167	(5) 47.1/12.96/364%
168	(6) 62.8/17.38/362%
169
170	(p)ibm-aix-regatta 700 9.5/9.7/98% 18.3/16.0/114%
171	(4) 38.3/24.7/155%
172	-----------------------------------------------------------------------------------
173
174	B.3/ Tests fft (FFTW , FFTPack )
175
176	(1) time cpupower 2
177	(2) time tfft 2000000 W d 0 0 (avec FFTW)
178	(3) time tfft 2000000 P d 0 0 (avec FFTPack_Sophya)
179
180	IndPerf=1000/TCPU
181
182
183	-----------------------------------------------------------------------------------
184	(1)MFLOPS (2)CPU/Elap/% (2)IndPerf (3)CPU/Elap/% (3)IndPerf
185	-----------------------------------------------------------------------------------
186	(a)xeon-lx-2.4GHz 1167 5.5/5.6/97% 180 7.4/7.4/100% 135
187	-O3 -g 6.8/7.1/96% 147
188	(b)xeon-lx-2.8GHz 920
189	(c)amd-lx 690 4.2/4.2/100% 238
190	~2x 4.7/4.7/99%
191
192	(d)osf-xp1000 648 9.9/10.2/97% 101 9.3/9.46/98.5% 107
193	(e)superosf 842 6./6.22/96.7% 166 5.1/5.18/97.4% 196
194	(f)osf-asc 420 13.3/13.9/95.7% 75 12.2/17.6/70% 82
195
196	(f)G4-osx-1.25GHz 333 15.2/15.9/96% 66 23.8/34.1/70% 42
197	(h)G5-osx-1GHz 1151 8.5/8.7/97.5% 120 11.5/11.6/99% 87
198	-tune=G5 5/5.1/98% 200 5.5/5.6/97% 180
199	(i)core-osx-1.83GHz 855 4.6/4.75/97% 217 7/7.08/99% 142
200	-O2 3.2/3.2/99% 312 3.6/3.6/99% 277
201	-O2 2 jobs // 2 x 3.9/4.3/90% 2x 4.7/5/91% 200
202	(j)xeon-osx 2600 2.6/2.6/98% 384
203	2 jobs // ~2x 3.6/4.1/87% 250
204	4 jobs //
205
206	(p)ibm-aix-regatta 700 6.25/18.9/33% 160 5.25/15.7/33% 190
207
208	-----------------------------------------------------------------------------------

Note: See TracBrowser for help on using the repository browser.

Download in other formats: