Active Topics
-
Maemo repository (5)
to Community by dredlok706 - 1 day, 10 hrs ago -
Re-commissioning N9 in 2018 (127)
to MeeGo / Harmattan by smatkovi - 1 day, 13 hrs ago -
Porting apps to Leste (41)
to Maemo 7 / Leste by Arno_11 - 2 days, 6 hrs ago -
List of well performing apps on Leste (33)
to Maemo 7 / Leste by teroyk - 2 days, 15 hrs ago -
N900 is not recognizable on any of my devices (windows 10,7, xp) What can I do? (1)
to Nokia N900 by nieldk - 3 days, 19 hrs ago -
Error after installing Leste on sdcard (13)
to Maemo 7 / Leste by Maemish - 6 days, 22 hrs ago - more...
 |
|
2011-03-01
, 04:38
|
|
Posts: 185 |
Thanked: 111 times |
Joined on Jul 2010
@ Mexico DF, Mexico
|
#22
|
My result with short script:
Kingston microSDHC 16GB Class 4
Read: 5.20291 MB/s
Write: 8.31255 MB/s
The slower Read I've seen around
I'm trying to decide if Buying Class 6 or Class 10, but I'm not convinced at all...
Kingston microSDHC 16GB Class 4
Read: 5.20291 MB/s
Write: 8.31255 MB/s
The slower Read I've seen around
I'm trying to decide if Buying Class 6 or Class 10, but I'm not convinced at all...
__________________
Linux Registered User # 492214
http://counter.li.org/
------------------------
N900 registered as Linux Machine # 426325
Linux Registered User # 492214
http://counter.li.org/
------------------------
N900 registered as Linux Machine # 426325
|
|
2011-03-01
, 05:14
|
|
Posts: 1 |
Thanked: 0 times |
Joined on Nov 2010
|
#23
|
Originally Posted by akih
How is it that your class 2 is faster than class 4 ?
I am create a MicroSD read/write speed benchmark script.
- Simple shell script
- No require additional software (busybox command only)
- Create 1GB file on MicroSD
sdbench.sh
usage:Code:#!/bin/sh MMC="/media/mmc1" DAT="$MMC/test.dat" CNT=1024 SDREAD="" SDWRITE="" grep -q "$MMC" /etc/mtab if [ $? == 0 ]; then echo "testing SD Card read/write speed. Please wait a few minutes." # Write test SDWRITE=`time -p dd if=/dev/zero of=$DAT bs=1M count=$CNT 2>&1|head -3|tail -1` # Read test SDREAD=`time -p cp $DAT /dev/null 2>&1|head -1` # Delete data file rm $DAT echo echo "Result" echo $SDREAD | awk '{print "Read:", 1024/$2, "MB/s"}' echo $SDWRITE | awk '{print "Write:", 1024/$2, "MB/s"}' else echo "SD Card not found." fi
- Copy sdbench.sh to home directory
- Open X-Terminal
- Run script (eg. sh sdbench.sh)
- wait a few minute
My result
SanDisc Class2 16GB
Read: 14.5952 MB/s
Write: 8.53831 MB/s
Buffalo Class4 8GB
Read: 13.4031 MB/s
Write: 4.84413 MB/s
|
|
2011-03-01
, 09:13
|
|
Posts: 1,258 |
Thanked: 672 times |
Joined on Mar 2009
|
#24
|
Sandisk is good :-)
Anyway, the class rating is a very specific measurement of /minimum/ speed in one very specific use, that resembles the kinda of load a card would see in a simple digital/video camera.
Your benchmarks are not measuring class rating, and neither class rating nor benchmarks measure how fast the card would perform for nitdroid or meego..
Actually it's probably impossible to determine the real proper class rating through benchmarks, it requires intimate knowledge of the internal structures and operation of the card in question, something only the manufacturer would know.
My own benchmarks of Sandisk, Kingston and Transcend cards have generally shown the sandisk class 2 as having an overall more consistent performance across a wide range of different type of loads, while class 6 kingston and transcend have beaten sandisk class 2 only in the sequential write/read patterns..
Anyway, the class rating is a very specific measurement of /minimum/ speed in one very specific use, that resembles the kinda of load a card would see in a simple digital/video camera.
Your benchmarks are not measuring class rating, and neither class rating nor benchmarks measure how fast the card would perform for nitdroid or meego..
Actually it's probably impossible to determine the real proper class rating through benchmarks, it requires intimate knowledge of the internal structures and operation of the card in question, something only the manufacturer would know.
My own benchmarks of Sandisk, Kingston and Transcend cards have generally shown the sandisk class 2 as having an overall more consistent performance across a wide range of different type of loads, while class 6 kingston and transcend have beaten sandisk class 2 only in the sequential write/read patterns..
 |
|
2011-03-01
, 10:16
|
|
Posts: 468 |
Thanked: 775 times |
Joined on May 2010
@ Hereford, England
|
#25
|
@one1002 - Since this thread has been bumped, I had another go, and following your instruction, I finally got it to work - After 15 minutes! The only card which I had enough space on for the test to be able to run, was a sandisk class 2. This is obviously why it took a long time, but 15 minutes.
Seems to me like a very long time to read and write, then delete just 1GB, but hey, got it to work, so thanks for your help and also thanks to akih for the script.
Seems to me like a very long time to read and write, then delete just 1GB, but hey, got it to work, so thanks for your help and also thanks to akih for the script.
|
|
2011-03-01
, 10:22
|
|
Posts: 1,680 |
Thanked: 3,685 times |
Joined on Jan 2011
|
#26
|
ahem, hdparm?
__________________
N900: One of God's own prototypes. A high-powered mutant of some kind never even considered for mass production. Too weird to live, and too rare to die.
| The Following User Says Thank You to vi_ For This Useful Post: | ||
|
|
2011-03-01
, 10:45
|
|
Posts: 99 |
Thanked: 65 times |
Joined on Jan 2008
@ Finland
|
#27
|
I know it really depends on what you want to test, but personally I prefer to divide speed tests to raw media speed and file system performance. For raw media speed read/write speed I use dd to read/write the media directly (e.g. dd if=random.static of=/dev/foo bs=1M). For file system performance, I've found bonnie++ to be sufficient. If you want to compile bonnie++ as a Debian package, just grab the sources from unstable; it should compile nicely for Maemo (at least for Diablo) as well.
|
|
2011-03-04
, 02:43
|
|
Posts: 1 |
Thanked: 0 times |
Joined on Mar 2011
|
#28
|
hello, please can you make a script for microSD detection issue?? I'm having problems with detection, mem card is ok I tried three, i converted the file system to ext2 as w2el but nit even dmesg shows anything... please help me
 |
|
2011-03-04
, 03:10
|
|
Posts: 566 |
Thanked: 282 times |
Joined on Sep 2010
@ Lower Saxony
|
#29
|
Result
Read: 12,5 MB/s
Write: 10 MB/s
Sandisk Premier 8GB from 2007. No Idea what class but they were the fastest cards on the market back then. Had lots of windows open when I ran the test. Might be faster right after boot.
Read: 12,5 MB/s
Write: 10 MB/s
Sandisk Premier 8GB from 2007. No Idea what class but they were the fastest cards on the market back then. Had lots of windows open when I ran the test. Might be faster right after boot.
|
|
2011-09-01
, 09:18
|
|
Posts: 561 |
Thanked: 75 times |
Joined on Jan 2010
@ Spain
|
#30
|
None of the scripts work for me.
He throws an error when using /dev /null.
The message reads:
I tried to test directly from the terminal and gives the same error. As much as "user" as "root"
I am testing a microSD 32GB Class 10 Lexor.
He throws an error when using /dev /null.
The message reads:
Code:
cp: write error: No space left on device Result: Read: awk: cmd. line: 1: Division by zero Write: awk: cmd. line: 1: Division by zero
I am testing a microSD 32GB Class 10 Lexor.
Read: 8.25764 MB/s
Write: 6.58328 MB/s