Start from nature and take a trip for venture

Wolfgang Spraul wolfgang at qi-hardware.com
Wed Jul 22 20:00:24 EDT 2009


Nicola,

I ran some tests on a NanoNote board yesterday, which has the Samsung
K9GAG08U0M NAND chip we will probably ship the Ben NanoNote with.
microSD card was a 16 GiB Silicon Power Class 6.

On the SD card:
  date; dd if=/dev/zero of=foo bs=1024k count=50; sync; date
    50 seconds = 1 MB / sec
  date; dd if=foo of=/dev/null bs=1024k count=50; sync; date
    13 seconds = 3.8 MB / sec

One the Samsung NAND chip (mounted via yaffs):
  date; dd if=/dev/zero of=foo bs=1024k count=50; sync; date
    21 seconds = 2.3 MB / sec
  date; dd if=foo of=/dev/null bs=1024k count=50; sync; date
    14 seconds = 3.5 MB / sec

This is all totally unoptimized or anything (2.6.24 kernel, the .28 kernel
wasn't working yesterday). I think we will first try to move everything to
more recent SW, .28 kernel or newer.
Wolfgang

On Thu, Jul 23, 2009 at 12:55:24AM +0200, Nicola Mfb wrote:
> On Thu, Jul 23, 2009 at 12:44 AM, W.G. van de Hulst<suyckerbol at gmail.com> wrote:
> >> Just my little wish. Please adopt a fast bus for flash/sd, yesterday I
> >> launched a dd if=/dev/zero of=foo bs=1024k count=10; sync on a friend
> >> iphone, and performed about 7.5 megabyte/s.
> >> This will contribute to responsiveness of the device when launching
> >> application.
> >
> > If you're lanching applications, don't you rather mean read speed?
> 
> Yes! thanks for pointing that, we made several tests and reported only
> the write one, as we meet again I'll report dd if=foo of=/dev/null too
> with a fresh restart to avoid buffer caching.
> 
> Regards
> 
>     Nicola
> 
> _______________________________________________
> Qi Developer Mailing List
> Mail to list (members only): developer at lists.qi-hardware.com
> Subscribe or Unsubscribe: http://lists.qi-hardware.com/cgi-bin/mailman/listinfo/developer




More information about the discussion mailing list


interactive