I get around 1600 kB/s with both a panasonic and a simpletech pro-x 256MB SD card when I dd 10MB from /dev/zero to /mnt/card. Is there a better way to measure the speed?
That\'s a good choice as /dev/zero is not going to be a bottle neck in terms of read. However 10mb may not be a large enough file for a true test.
You have 64mb ram, some of that will be used as buffer cache so to defeat the effects of that you may have to write a larger file.
Just to prove the point, post the exact command you ran and I will try it on my two sd cards. Then the only difference would be the make of sd card in our test.
Peter
The following is with a vfat-formatted 256MB Simpletech Pro-X SD card. I include a sync command into the timing to exclude buffer effects.
root@zaurus(pts0):~# time sh -c \'dd if=/dev/zero of=/mnt/card/dump bs=1024k count=10; sync\'
10+0 records in
10+0 records out
real 0m5.942s
user 0m0.060s
sys 0m5.880s
root@zaurus(pts0):~# rm -f /mnt/card/dump
root@zaurus(pts0):~# time sh -c \'dd if=/dev/zero of=/mnt/card/dump bs=1024k count=20; sync\'
20+0 records in
20+0 records out
real 0m11.741s
user 0m0.040s
sys 0m11.680s
root@zaurus(pts0):~# rm -f /mnt/card/dump
root@zaurus(pts0):~# time sh -c \'dd if=/dev/zero of=/mnt/card/dump bs=1024k count=40; sync\'
40+0 records in
40+0 records out
real 0m23.596s
user 0m0.040s
sys 0m8.430s
root@zaurus(pts0):~# rm -f /mnt/card/dump
root@zaurus(pts0):~# time sh -c \'dd if=/dev/zero of=/mnt/card/dump bs=1024k count=80; sync\'
80+0 records in
80+0 records out
real 0m47.895s
user 0m0.010s
sys 0m26.370s
Not sure why the sys time is too small for the larger file sizes. The real time is correct. These values are reproducible within a few percent.