Active Topics
-
Is there a section to talk about Java ME? Apps, etc. (7)
to General by teroyk - 7 hrs, 57 mins ago - more...
|
2010-04-11
, 06:42
|
|
Posts: 1,751 |
Thanked: 844 times |
Joined on Feb 2010
@ Sweden
|
#12
|
My fault.. forgot the extra nbench.. have edit it now.
chmod 777 /home/opt/nbench/nbench
chmod 777 /home/opt/nbench/nbench
|
|
2010-04-11
, 07:42
|
|
Posts: 1,751 |
Thanked: 844 times |
Joined on Feb 2010
@ Sweden
|
#13
|
Did you get it Jakiman?
|
|
2010-04-11
, 08:03
|
|
Posts: 356 |
Thanked: 172 times |
Joined on Jan 2010
@ Canada
|
#14
|
Originally Posted by craftyguy
Sorry -- that's a holdover from my PC benchmarking days, and probably doesn't really matter very much when applied to the N900.
Just curious, why would you want to disable sound? Unless it's not very common to play the game with sound, you'd get a much more rounded stress test (of all system components) if you ran a benchmark of a game with the same settings as a normal play session.
It's like doing time trials on a racetrack with racing tires, and then swapping them out for your street tires on the day of the race :P
The theory was that different sound cards/drivers could affect the benchmark in unpredictable ways (maybe because of sound being synced to video in some cases ??), and so leaving it enabled would cause the demo to play back in a way that wouldn't accurately reflect the capabilities of the CPU/GPU, which is what people were generally most interested in measuring.
On a closed device like this one, that's probably not much of a concern, but I decided to list the command anyway just in case. You know.. in the interest of science
 |
|
2010-04-11
, 08:35
|
|
Posts: 168 |
Thanked: 206 times |
Joined on Apr 2010
@ Finland
|
#15
|
Here are couple of Nbench tests with Titan's uvl kernel:
(Linux 2.6.28.10maemo-ulv-omap1)
Running @850MHz (maemo20)
Running @900MHz (maemo20)
Running @1000MHz (maemo21)
Did not completed test @1100MHz (maemo21)
850MHz is default for normal usage
Last edited by Gusse; 2010-04-11 at 13:55. Reason: More tests added
(Linux 2.6.28.10maemo-ulv-omap1)
Running @850MHz (maemo20)
Running @900MHz (maemo20)
Running @1000MHz (maemo21)
Did not completed test @1100MHz (maemo21)
850MHz is default for normal usage
Last edited by Gusse; 2010-04-11 at 13:55. Reason: More tests added
 |
|
2010-04-11
, 09:04
|
|
Posts: 664 |
Thanked: 160 times |
Joined on Jul 2008
@ Australia
|
#16
|
I have the following with TMO and WiFi on:
BTW I'm using titan's kernel-maemo20 with 810MHz
Code:
BYTEmark* Native Mode Benchmark ver. 2 (10/95)
Index-split by Andrew D. Balsa (11/97)
Linux/Unix* port by Uwe F. Mayer (12/96,11/97)
TEST : Iterations/sec. : Old Index : New Index
: : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT : 333.6 : 8.56 : 2.81
STRING SORT : 58.633 : 26.20 : 4.06
BITFIELD : 1.1276e+08 : 19.34 : 4.04
FP EMULATION : 83.333 : 39.99 : 9.23
FOURIER : 1562.6 : 1.78 : 1.00
ASSIGNMENT : 5.3861 : 20.49 : 5.32
IDEA : 899.71 : 13.76 : 4.09
HUFFMAN : 525.29 : 14.57 : 4.65
NEURAL NET : 1.6411 : 2.64 : 1.11
LU DECOMPOSITION : 60.911 : 3.16 : 2.28
==========================ORIGINAL BYTEMARK RESULTS====================
INTEGER INDEX : 18.393
FLOATING-POINT INDEX: 2.454
Baseline (MSDOS*) : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW======================
CPU :
L2 Cache :
OS : Linux 2.6.28.10maemo-ulv-omap1
C compiler : gcc version 4.2.1
libc :
MEMORY INDEX : 4.433
INTEGER INDEX : 4.711
FLOATING-POINT INDEX: 1.361
Baseline (LINUX) : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38
* Trademarks are property of their respective holder.
|
|
2010-04-11
, 09:59
|
|
Posts: 1,427 |
Thanked: 2,077 times |
Joined on Aug 2009
@ Sydney
|
#17
|
thanks almehdi. got it working! 
Here's mine using Lehto's ulv kernels at 1000mhz and 1100mhz. (2 different kernels)
@800mhz ulv (Lehto)
@1000mhz ulv (Lehto)
@1100mhz ulv (Lehto)
Last edited by jakiman; 2010-04-11 at 11:38.
Here's mine using Lehto's ulv kernels at 1000mhz and 1100mhz. (2 different kernels)
@800mhz ulv (Lehto)
TEST : Iterations/sec. : Old Index : New Index
: : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT : 328.92 : 8.44 : 2.77
STRING SORT : 57.537 : 25.71 : 3.98
BITFIELD : 1.1172e+08 : 19.16 : 4.00
FP EMULATION : 82.054 : 39.37 : 9.09
FOURIER : 1549.6 : 1.76 : 0.99
ASSIGNMENT : 5.3297 : 20.28 : 5.26
IDEA : 889.68 : 13.61 : 4.04
HUFFMAN : 510.16 : 14.15 : 4.52
NEURAL NET : 1.6193 : 2.60 : 1.09
LU DECOMPOSITION : 60.568 : 3.14 : 2.27
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX : 18.110
FLOATING-POINT INDEX: 2.432
Baseline (MSDOS*) : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU :
L2 Cache :
OS : Linux 2.6.28-omap1
C compiler : gcc version 4.2.1
libc :
MEMORY INDEX : 4.376
INTEGER INDEX : 4.630
FLOATING-POINT INDEX: 1.349
Baseline (LINUX) : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38
@1000mhz ulv (Lehto)
TEST : Iterations/sec. : Old Index : New Index
: : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT : 412.16 : 10.57 : 3.47
STRING SORT : 72.171 : 32.25 : 4.99
BITFIELD : 1.3954e+08 : 23.94 : 5.00
FP EMULATION : 102.6 : 49.23 : 11.36
FOURIER : 1937 : 2.20 : 1.24
ASSIGNMENT : 6.6563 : 25.33 : 6.57
IDEA : 1113.3 : 17.03 : 5.06
HUFFMAN : 648.84 : 17.99 : 5.75
NEURAL NET : 2.0258 : 3.25 : 1.37
LU DECOMPOSITION : 75.73 : 3.92 : 2.83
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX : 22.712
FLOATING-POINT INDEX: 3.041
Baseline (MSDOS*) : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU :
L2 Cache :
OS : Linux 2.6.28-omap1
C compiler : gcc version 4.2.1
libc :
MEMORY INDEX : 5.473
INTEGER INDEX : 5.818
FLOATING-POINT INDEX: 1.687
Baseline (LINUX) : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38
TEST : Iterations/sec. : Old Index : New Index
: : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT : 453.36 : 11.63 : 3.82
STRING SORT : 79.168 : 35.37 : 5.48
BITFIELD : 1.5369e+08 : 26.36 : 5.51
FP EMULATION : 112.88 : 54.17 : 12.50
FOURIER : 2127.9 : 2.42 : 1.36
ASSIGNMENT : 7.3384 : 27.92 : 7.24
IDEA : 1223.3 : 18.71 : 5.56
HUFFMAN : 712.88 : 19.77 : 6.31
NEURAL NET : 2.2239 : 3.57 : 1.50
LU DECOMPOSITION : 83.014 : 4.30 : 3.11
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX : 24.978
FLOATING-POINT INDEX: 3.338
Baseline (MSDOS*) : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU :
L2 Cache :
OS : Linux 2.6.28-omap1
C compiler : gcc version 4.2.1
libc :
MEMORY INDEX : 6.022
INTEGER INDEX : 6.396
FLOATING-POINT INDEX: 1.851
Baseline (LINUX) : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38
__________________
JAKIMAN's ShootSpeak.com Blog Follow me on twitter
JAKIMAN's N900 Performance Tweaks / Overclocking Article
JAKIMAN's Useful tips / hacks / tricks / apps for Nokia N9
JAKIMAN's ShootSpeak.com Blog Follow me on twitter
JAKIMAN's N900 Performance Tweaks / Overclocking Article
JAKIMAN's Useful tips / hacks / tricks / apps for Nokia N9
Last edited by jakiman; 2010-04-11 at 11:38.
| The Following 2 Users Say Thank You to jakiman For This Useful Post: | ||
|
|
2010-04-11
, 10:58
|
|
Posts: 13 |
Thanked: 10 times |
Joined on Apr 2010
@ Stockholm
|
#18
|
This are my nbench result running with a stock kernel.
PHP Code:
TEST : Iterations/sec. : Old Index : New Index
: : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT : 247.36 : 6.34 : 2.08
STRING SORT : 43.575 : 19.47 : 3.01
BITFIELD : 8.3648e+07 : 14.35 : 3.00
FP EMULATION : 61.886 : 29.70 : 6.85
FOURIER : 1164.7 : 1.32 : 0.74
ASSIGNMENT : 3.9968 : 15.21 : 3.94
IDEA : 667.19 : 10.20 : 3.03
HUFFMAN : 389.11 : 10.79 : 3.45
NEURAL NET : 1.2144 : 1.95 : 0.82
LU DECOMPOSITION : 44.823 : 2.32 : 1.68
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX : 13.646
FLOATING-POINT INDEX: 1.817
Baseline (MSDOS*) : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU :
L2 Cache :
OS : Linux 2.6.28-omap1
C compiler : gcc version 4.2.1
libc :
MEMORY INDEX : 3.291
INTEGER INDEX : 3.494
FLOATING-POINT INDEX: 1.008
| The Following User Says Thank You to _privateMember For This Useful Post: | ||
|
|
2010-04-11
, 11:09
|
|
Posts: 664 |
Thanked: 160 times |
Joined on Jul 2008
@ Australia
|
#19
|
Actually., so how do one figure out whose is running as per the MHz? and is it doing any performance boost as per this benchmark?
|
|
2010-04-11
, 11:10
|
|
Posts: 170 |
Thanked: 75 times |
Joined on Jun 2008
@ NYC
|
#20
|
Cortex A8@900mhz using Titan's ulv.
Code:
BYTEmark* Native Mode Benchmark ver. 2 (10/95)
Index-split by Andrew D. Balsa (11/97)
Linux/Unix* port by Uwe F. Mayer (12/96,11/97)
TEST : Iterations/sec. : Old Index : New Index
: : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT : 370.4 : 9.50 : 3.12
STRING SORT : 65.601 : 29.31 : 4.54
BITFIELD : 1.2545e+08 : 21.52 : 4.49
FP EMULATION : 92.44 : 44.36 : 10.24
FOURIER : 1745.6 : 1.99 : 1.12
ASSIGNMENT : 5.9572 : 22.67 : 5.88
IDEA : 1002.4 : 15.33 : 4.55
HUFFMAN : 583.43 : 16.18 : 5.17
NEURAL NET : 1.8175 : 2.92 : 1.23
LU DECOMPOSITION : 68.358 : 3.54 : 2.56
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX : 20.447
FLOATING-POINT INDEX: 2.738
Baseline (MSDOS*) : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU :
L2 Cache :
OS : Linux 2.6.28.10maemo-lv-omap1
C compiler : gcc version 4.2.1
libc :
MEMORY INDEX : 4.931
INTEGER INDEX : 5.235
FLOATING-POINT INDEX: 1.519
Baseline (LINUX) : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38
* Trademarks are property of their respective holder.
It's like doing time trials on a racetrack with racing tires, and then swapping them out for your street tires on the day of the race :P