-
Notifications
You must be signed in to change notification settings - Fork 2
/
benchmark2
73 lines (69 loc) · 2.85 KB
/
benchmark2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
; benchmark for gifa
verbose 1 ; check what you're doing
; benchmark for gifa
; run on a 512 x 2k real data set.
; Processing performed is :
; - Exponen. Broading in F2
; - Real FT, phasing, in F2
; - 5 points Spline base-line correction in F2
; - cosine apodization in F1, zero-filling in F1
; - Real FT and phasing in F1.
; The displayed spectrum at the end is 512 by 1k real points
; results are (some figures date from version 1.0 !) :
; * 3'50" - Dec VS 3100 VMS 5.2 * 1'30" - SUN Sparc 330
; * 2'30" - Macintosh Centris 650 - running MachTen
; * 53" - Silicon-Graphics 4D/20 * 47" - S.G. 4D/25
; * 37" - SUN SparcStation 2 * 34" - DEC Station 5000-200
; * 36.3" - PC Am486DX4/100 Linux
; * 27" - S.G. 4D/30 * 31" on IBM Risc 6000 320
; * 20" - Mac 7100 80MHz running MkLinux
; * 20" - Alliant FX 40 3CE * 17.5" - SUN Sparc-10/30
; * 17" - HP 725/50 * 16" - Indigo R4000
; * 15.5" - INDY PC * 15" - IBM Risc6000 550
; * 12.6" - IBM Risc6000 560 * 12.3" - SUN Sparc-20
; * 11.7" - INDY R4600 132MHz
; * 10.7" - SUN SparcCenter 2000E SuperSparc 85 Mhz
; * 10.5" - INDY R4400 174MHz * 10" - HP 715/80
; * 9.5" - PC Pentium 133 Linux * 7.5" - HP 735/99
; * 7.4" - HP PA-8000 160MHz * 7.0" - Mac 8600 200MHz
; * 6.8" - INDY R5000 180MHz
; * 6.5" - SGI-O2 R5000-PC * 5.06" - Pentium Pro 200
; * 4.6" - SGI-O2 R12000 270MHZ (compiled in 32 bits)
; * 4.16 - SGI-O2 R5000 300MHz
; * 3.8" - iMac 266MHz running linuxPPC
; * 3.3" - SGI Origin 200 R10000 180MHz
; * 3.25" - Pentium II 300MHz
; * 3.2" - SUN/Entreprise CPU UltraSparcII 250MHz
; * 2.67" - SGI Origin 200 R10000 225MHZ
; * 2.5" - PowerBook G3 400MHz running MacOsX
; * 2.49" - SGI Octane R10000 250MHz
; * 2.3" - Mac G3 400MHz LinuxPPC 2000
; * 2.13" - Pentium II 450MHz Linux RH 6.0
; * 2.0" - Mac G4 400MHz running MacOsX
; * 1.60" - Mac G4 800 MHz MacOsX (still to be optimized)
; * 1.31" - Pentium III 800MHz Linux RH 6.2
; * 1.20" - Pentium III 800MHz Linux Mandrake 8.0
; * 0.94" - Pentium IV 1.5GHz Linux RH 7.1
; first prepare a pseudo-data set for benchmarking GIFA
; set-up a 512 x 2k real data set
dim 2
chsize 128 128 one sin 0 f12 revf f1 revf f2 ; put something in it
chsize 512 2k itype 0 ; and zero-fill
put data
sh date
timer 0
timer 1
for i = 1 to 10
get data
em 0 1 revf f2 rft f2 ; main F2 processing
phase 20 20 f2 real f2 ;throwing Imaginary parts is needed here
bcorr 2 2 f2 10 100 300 600 900 0 ; for baseline corr.
chsize % (%+2)
sin 0 f1 chsize (%*2) % rft f1 ; main F1 proc
phase 20 20 f1 ;done
endfor
;sh 'sleep 3'
verbose 0
sh date
print 'Divide timer results by 10'
exit