On Thu, Dec 11, 2008 at 10:02:10PM -0800, Keith Lofstrom wrote:
> Speed testing CF cards with ALIX
Version 2 with corrections
Speed testing CF cards with ALIX
--------------------------------
Summary - it doesn't make much difference what you use.
I tried 3 different recently purchased 2GB CF cards for
the ALIX. I did 5 experiments, one trial each. I would
need to do more trials to find average differences that
I could reliably differentiate - nothing was spectacularly
better or worse. I set up a 200MB partition (of which I
am using 3%).
test 1 - initial write of the card, with a Digital Concepts USB
CF programmer. Timed by "dd" output. Note - this interface
came from Radio Shack, and melted (!!!) a few minutes ago,
so I can't re-run the Kingston test. I also have a PCMCIA
to CF adapter, and that programs a card in about 60 seconds
instead of 100. I don't often use it because I don't like
wearing out the PCMCIA slots in my laptop if I don't have to ...
test 2 - first boot. This takes a little extra time, as it
zeros blocks in the JFFS partition. I timed the boot manually
from first power to last message.
test 3 - second boot. Time from "reboot" command to last
message, with initial contents.
test 4 - third boot. Same as test 3, except I scp'ed a bunch
of overlay files and additional stuff into the jffs area.
test 5 - UDMA boot. I set the ALIX board to UDMA mode, to
see if that speeds up the card reading (hint - it doesn't)
The cards:
Kingston generic 2GB from Newegg. This is Kingston's basic
card. Kingston claims other versions are faster.
SanDisk Ultra II "15MB/sec" supposedly faster 2GB. There
is an Ultra III which is supposedly faster. 4x the cost
of the Kingston
Transcend 266x from Newegg. This is Transcend's "fast"
line, and about 50% more expensive. Very extensive
data sheet, specific electrical and timing and software
details, suitable for board-level and kernel driver
design.
Results (all times in seconds):
Card Write Boot Boot Boot Boot
test1 test2 test3 test4 test5
Kingston 72? 151 41 46 38
SanDisk 100 160 38 45 39
Transcend 101 151 43 41 42
Conclusion: while there appear to be small differences, they
would probably average out with repeated tests. The speeds
are probably limited by something besides the CF card, such
as DHCP server times, other activities on the computer with
the CF card write on it, etc. I would like to try this out
with a really old camera card (256MB or larger), but I
suspect it doesn't matter a whole lot what gets used;
reliability is probably more important than speed.
Keith
--
Keith Lofstrom [email protected] Voice (503)-520-1993
KLIC --- Keith Lofstrom Integrated Circuits --- "Your Ideas in Silicon"
Design Contracting in Bipolar and CMOS - Analog, Digital, and Scan ICs
--~--~---------~--~----~------------~-------~--~----~
The Personal Telco Project - http://www.personaltelco.net/
Donate to PTP: http://www.personaltelco.net/donate
Archives: http://news.gmane.org/gmane.network.wireless.portland.general/
Etiquette: http://www.personaltelco.net/index.cgi/MailingListEtiquette
List information: http://lists.personaltelco.net
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
-~----------~----~----~----~------~----~------~--~---
