I have a SanDisk Extreme III 2GB (at least I think it's III) in my
Soekris net5501 :

[EMAIL PROTECTED] $ dd if=/dev/zero of=nulls bs=65536 count=1600
1600+0 records in
1600+0 records out
104857600 bytes transferred in 8.604 secs (12186647 bytes/sec)
[EMAIL PROTECTED] $ dd if=nulls of=/dev/null bs=65536
1600+0 records in
1600+0 records out
104857600 bytes transferred in 9.118 secs (11499110 bytes/sec)
[EMAIL PROTECTED] $ dmesg | grep SanDisk
wd0 at pciide0 channel 0 drive 0: <SanDisk SDCFX3-2048>
[EMAIL PROTECTED] $ uname -a
OpenBSD tuna.zrh.weirdnet.ch 4.2 GENERIC#452 i386
[EMAIL PROTECTED] $ sysctl hw
hw.machine=i386
hw.model=Geode(TM) Integrated Processor by AMD PCS ("AuthenticAMD" 586-class)
hw.ncpu=1
...

It's pretty fast, especially considering it's CF on not the worlds
fastest machine.

Cheers,

Paul 'WEiRD' de Weerd


On Fri, Nov 16, 2007 at 04:02:32PM +0000, Christian Weisgerber wrote:
| Stefan Klein <[EMAIL PROTECTED]> wrote:
| 
| > If I use exactly the same commands / mount options as you I get less than 
| > 1MB/s....
| > 
| > I know that I cannot expect a "good" performance with the CF card, but 
5MB/s 
| > would just be fine :-)
| 
| I've started playing with a CF in a bigger machine, and the results
| are rather disappointing.
| 
| wd1 at pciide2 channel 0 drive 0: <TRANSCEND>
| wd1: 1-sector PIO, LBA, 7775MB, 15924384 sectors
| wd1(pciide2:0:0): using PIO mode 4, Ultra-DMA mode 4
| 
| This is a new "266x" CF.  Write performance is in the 200..300 kB/s
| range.  It takes six hours or so to copy the OpenBSD CVS repository
| on that drive.  Sequential read performance tops out at about
| 2300 kB/s.  Random access read throughput--running CVS checkouts
| in parallel, local and NFS--ends at ~1750 kB/s.
| 
| Interestingly, the older and supposedly slower CF in my Soekris 5501
| 
| wd0 at pciide0 channel 0 drive 0: <SanDisk SDCFH-1024>
| wd0: 4-sector PIO, LBA, 977MB, 2001888 sectors
| wd0(pciide0:0:0): using PIO mode 4, DMA mode 2
| 
| reads sequentially at just short of 5 MB/s.
| 
| I suspect the main performance factor to be the N-sector capability.
| The Transcend generates an interrupt for each 512-byte sector (which
| eats a ridiculous share of CPU on my Opteron box), the SanDisk a
| fourth of that.
| 
| IIRC, the SanDisk above is from their "Ultra II" line.  I wonder
| how the newer "Extreme III" and "Extreme IV" perform.
| 
| -- 
| Christian "naddy" Weisgerber                          [EMAIL PROTECTED]
| 

-- 
>++++++++[<++++++++++>-]<+++++++.>+++[<------>-]<.>+++[<+
+++++++++++>-]<.>++[<------------>-]<+.--------------.[-]
                 http://www.weirdnet.nl/                 

Reply via email to